Why Super-Fast Internet Is Coming Super Slowly | Hacker News Reader

Why Super-Fast Internet Is Coming Super Slowly

online.wsj.com

129 pointsbocalogic124 comments

I view the broadband industry as a fascinating example of what happens when an infrastructure industry meets the technology industry. TWC and friends were established to enact a business plan that's been viable and profitable for centuries: laying down expensive infrastructure is economically feasible if most of the cost comes up front and you can charge rents on it for decades. Stability is crucial. This is the economic reasoning that built the cable TV network, and the phone network before that, and the power grid before that, and the railroad network before that, etc.

Trouble is, this business model severely discourages new development (witness the "No one wants gigabit" rationale). Why interrupt your twenty year plan just to extend its life a year or two into the future? This isn't a bad thing if your industry is basically a utility industry, where technological changes move at a glacial pace. God help you if you enter into the technology industry armed only with the economic and managerial tools of a utility juggernaut.

So we've got ourselves a conflict. On one side there are irritated customers wondering why their service is so poor and why their bills are so high. On the other there are communications utilities providers: while they have a legitimate business need to keep their networks profitable, they are paying the price for committing the gravest sin possible in the technology world: complacency.

So why does broadband in other countries not have this problem, like Scandinavia, or Korea & Japan?

For the same reasons those countries implement things like universal health care. The crux of the issue with the US however is that being powerful here can mean greater power elsewhere. For example, if I have 100% market share in Sweden, it doesn't necessarily give me the capital and muscle to become the biggest internet provider in all of Europe. Whereas if you become the superpower in the US, expansion (read -> growth) is much easier to obtain.

Sprawl makes infrastructure investment orders of magnitude more expensive. For example, a small business [1] provides the fastest consumer access in the city of Seattle, but they refuse to do business outside of extremely high density buildings. There is no way a small business could do the same in McSuburbia. In the sprawly world outside of the US's dense cities, only municipal governments with a lot of political willpower have the resources to invest in the infrastructure needed for fast internet.

[1] http://www.condointernet.net/

I don't have direct experience with the state of the internet in the U.S., but isn't it terrible in large, metropolitan cities as well? I'm really just using the anecdata I've been reading on HN.

Furthermore, afaik many small towns even in eastern europe(while making great strides, no particularly huge cities) boast 100 mbit connections.

Assuming both of those statements are true, It doesn't seem to me like the economies of scale argument is valid here.

EDIT: 100 mbit/$20, 500 mbit/$30, checked my ISP.

You're absolutely correct. Population density is a red herring in this issue. The broadband situation is no better in the major cities of the U.S than it is in the rest of the country. Major cities in the rest of the developed world have much better connections somehow.

One of the first deployments of gigabit fiber in the U.S. was in Chatanooga TN, with a population density of 471.9/km2. New York City has a population density of 10,640/km2. Romania's third largest city, Timișoara, with population density of only 2,446.58/km2 is considered the city with the fastest internet in the world.

http://www.romania-insider.com/romanian-city-comes-out-first...

Density is not something that can be described adequately in a single number. It is entirely possible to have high residential densities in cities with remarkably low average densities. Moscow is a great example of this...their plazas and parks drive down average density way beyond the density that the average person experiences.

Everybody talks about the last mile being the most expensive part of infrastructure, but if the last meter is 90% of the cost of the last mile (I don't know what the exact proportion is, but it is definitely close), then 100 houses in 10 sq km census block is going to be in a cost category that is orders of magnitude higher than 100 apartments in a tower with a 9.9 sq km park around it.

Density either prevents better infrastructure in the US, or it doesn't. My use of statistics was merely to show that density has no relation to the existence of fiber internet infrastructure. The US cannot be too sprawling for fiber if areas with lower density in other countries have better internet. Major cities cannot also be too dense when cities with higher densities abroad also have better internet.

This is a commonly repeated argument in favor of the status-quo in the US, and I'm just trying to refute it whenever possible.

Your use of statistics is meaningless because average density is meaningless. It suffers from the same resolution problem as the coastline paradox [1]. I'm an example of this: If you measure density by the block, I live in a density of 120,000 persons/km2, but if you measure density by the city, I live in a density of 1720 persons/km2, and if you measure density by the MSA, I live in a density of 168 persons/km2. I have gigabit, and the fact that I live in an MSA with 168 persons/km2 is in no way a refutation of the fact that density matters for infrastructure costs.

[1] http://en.wikipedia.org/wiki/Coastline_paradox

Saying average density is meaningless is like saying, because of the coastline paradox, even a small island's coastline is the same length as a large continent's coastline -- they're both infinite!

Correct. However, if you can define a level of resolution to measure the coastline/density, then then it can be compared. Average density/coastline is meaningless without context.

Great point.

That still doesn't explain why there's no gigabit FTTH in NYC, with its high average density.

It sure as heck doesn't explain why there's no gigabit FTTH in Manhattan proper--by your reckoning the borough should have a residential density even higher than its overall average would suggest by virtue of Central Park taking up so much of the upper half.

High density doesn't eliminate millions of other reasons that a given city doesn't have it. It just provides a floor for cost competitiveness, and a major input into ROI. Even in Kansas City, google has stated they won't roll out FTTH to every neighborhood, even in the long term...and in the short term, they have focused on the neighborhoods with the fastest ROI. I'll let you guess whether the high density or the low density gets built first.

Then why did you bring up density in the first place?

Unless you're saying that urban and suburban America each lack FTTH for their own wholly unrelated reasons, an explanation that seems rather lacking in parsimony.

[1] http://en.wikipedia.org/wiki/Occam%27s_razor

Because a huge and significant chunk of America lives at a density that is prohibitive for small competitors to provide FTTH. The only way to get it is through municipal funding, because the large broadband providers certainly aren't going to provide it.

>> I don't have direct experience with the state of the internet in the U.S., but isn't it terrible in large, metropolitan cities as well?

It depends on where you are. Two different apartment buildings in DC or NYC could differn, one getting FIOS and the other on a congested cable network or low-speed phone company DSL circuit.

But don't get the idea that the internet is generally terrible in the USA. I'd imagine my situation is similar to most others - I live in a smaller town (population < 50,000) and about 90 miles from the nearest 'big' city (Detroit), but for $39.99 I get 30mb cable internet which I'm more than happy with. They offer 60mb and 100mb packages but I don't see the sense in spending any more.

That doesn't explain why internet speeds suck in NYC.

Population density in NYC and SFO actually hurt new rollouts. That is why those places and Los Angeles will probably be dead last on a Google Fiber rollout (if ever).

The old NYC infrastructure and all of the physical work (and permits, contracts, everything else) means a huge overhaul would most likely not be worth doing in dense and already wired places like these.

Also rayiner pointed out in another thread that according to Akamai the US already is pretty high up in the list of overall bandwidth. Also I think the HN / Reddit / tech echo chamber overplays the demand for gigabit internet. Most of America has no need for this, nor the willingness to pay for a connection this fast. 10-25 Mb is fine for most people. Browse Facebook, surf the web, stream a couple of movies, play games.

I would be willing to pay for a gigabit connection if the prices were lower. In Japan and South Korea for instance, a consumer can get a gigabit connection for the equivalent of about 40USD. I currently pay more for a connection 1/20th the speed. If their telcos can do it, ours can too. However ours want us to pay more (much, much more) for gigabit, which is why there is the argument that Americans don't want to pay for gigabit. Yeah, charge a reasonable price and you'll have more demand for gigabit speeds.

Also: I know my dissent in pointing out that gigabit internet for $30/month everywhere in the USA isn't going to magically happen isn't popular around here, but you shouldn't down vote something you disagree with.

You're being downvoted because you're wrong. Density is a red herring that can be easily demonstrated by looking at comparative examples.

A single city in another country is apples and oranges, it is wrong to assume that since a single city in another country can do it, then it should be simple to implement across the USA or in cities which are wholly different politically in the US.

Look, you can't have it both ways. Either population density helps or hinders a fiber rollout--if it helps, then why is the urban U.S.A. just as far behind as the rest? If it hinders, then why do HK and SK have it when NYC doesn't?

As for the remainder--well of course they're not making use of a service that they have no way to buy in the first place...and your insinuations remind me of the people in the early 19th century who howled that the first locomotives' 10 mph crusing speed was far faster than any sane person would ever want to travel.

The urbanized parts of the U.S. aren't behind in general. If the east coast states were countries, a number would be top-5 on Akamai's list for highest average bandwidth. As for HK and SK: there is only one city in the U.S. as dense as Hong Kong or Seoul, and that's NYC. And it has FIOS. So why does everywhere in NYC not have FIOS? Is it because of lobbying? No, its because building infrastructure in dense places requires coordination, and its just not a big political priority the way it is in Asian countries. Its why infrastructure projects of all kinds cost more in the U.S. from trains to telecom. In other countries, the government makes it a priority to remove roadblocks to infrastructure development, in a top down way. In the U.S., every infrastructure project becomes mired in NIMBYISM. Look in the back yard of Silicon Valley. Why isn't SF wired with gigabit. Is it because Comcast had more lobbying muscle than tech in tech's own back yard? Its an absurd proposition. The roadblock is instead the same dysfunctional government and the voters that complain about Google busses.

I must say your dichotomy between lobbying and political priorities is positively Martian.

And I haven't noticed the U.S. public rallying in horrified opposition to faster internet speeds in a big way. Or at all, really.

They haven't been rallying in support of it either.

You attributed the slow broadband rollout to public hostility, not public indifference.

Where did I say the public was hostile to faster Internet? The public is for it, just that this group is too small for them to be relevant yet.

Its a combination of density, economics, and politics. Japan/Korea are best case scenario. Not only very dense, but people believe in tech as a way forward economically. In the U.S., the places that are dense also have intractable politics. Outside the tech set, you're not going to sell San Franciscans on the idea that gigabit should be an important priority. They couldn't care less.

> Japan/Korea are best case scenario. Not only very dense, but people believe in tech as a way forward economically.

This is a common misconception. When it comes to actually using technology, the Japanese are at least a decade behind the West, thanks to their aging, collectivist, and risk-averse society. Fax machines continue to be widely used[0].

The primary difference is that in Japan, the government nationalized the last mile, resulting in a ton of competition. You can pick between a dozen ISPs in Tokyo.

0: http://www.nytimes.com/2013/02/14/world/asia/in-japan-the-fa...

My internet speeds are pretty spiffy here in NYC; granted, not fiber spiffy.

Of course, I'm with Cablevision - when most of the city is limited to TWC.

Population density is a complete red herring in this context. It is the excuse that telecom and cable companies rely on for their lack of investment. See my comment below for the hard numbers of how this is simply not true.

Sweden here: 15 years ago we were happily renting a 1.5Mbps internet connection in a very small village in the desolate areas of northern sweden. The network was fiber though and was quickly upgraded to 10 and 100 Mbps. Cost was about $60 per month in the start, now it is $38 per month. No limits the last 10 years.

Take Norway; here all the copper lines used for DSL used to be owned by the national phone company. Then competition was established by forcing the national phone company to sell access to the lines at a set rate to anyone that wanted to deliver internet access. A monthly "line rental" is charged to the end customer - so that the cost is transparent. "Line rental" is about $13/month and a 20Mb/1Mb ADSL is $50/month (Cheapest 100Mb/100Mb (fiber) is $80.) There is a site for comparing prices (telepriser.no) run by the equivalent to the FCC, showing all the suppliers offerings.

> I view the broadband industry as a fascinating example of what happens when an infrastructure industry meets the technology industry.

I don't disagree with your comment, but your argument would seem to apply to the rest of the world, whereas in fact the US is in a fairly unique situation among developed countries.

Most of the current situation can be explained by the fact that broadband is currently the only unregulated public utility.

In NYC, I don't really pay much attention to electricity, gas, and water. ConEd kind of sucks, and I wish they were better, but really, they work well enough most of the time, and I'm not sure that I could name anybody else who would do it better.

The problem is that we sold off monopoly rights to broadband without attaching any strings of regulation, etc. Why should they bother to upgrade their infrastructure? Heck, we literally paid the cable companies billions of dollars of government money to do exactly that, and because we failed to enforce that with any real teeth, they just took the money and ran. Of course they're not going to listen when consumers say they "want gigabit". It would make no business sense to do otherwise, especially when 80% of your customers are unable to vote with their wallet and use a competing service[0].

There's a case to be made for either a free-market approach to broadband, or a more "socialized", regulated monopoly. Unfortunately, what we currently have is a hybrid of both, and it happens to hit upon the worst of both worlds.

[0] There's a South Park clip that does a great job of describing this relationship. I won't link to it here, because it's rather crude and possibly (?) a bit NSFW, but searching "south park cable" on Youtube should lead you to it if you want to laugh, sigh, and cry all at the same time.

I disagree in terms of regulation: there is plenty of regulation on broadband internet, except it tends to discourage competition against broadband providers.

This isn't a problem that is going to be solved by placing regulations on the industry. Your example of past subsidies provides an excellent argument for this: given the political clout these entities hold and have in the past used to stilt legislation in their favor, what leads us to believe we're going to do any better with new regulations?

At the end of the day, the market is demanding better internet service. While telecommunications providers are certain to wage war against the injection against honest competition into their sector, things are actually looking pretty good in my mind. The FCC sees market demand and is working to tear down anti-competitive legislative barriers [1]. Smaller efforts such as Google Fiber are doing the hard work of convincing the public that better service is within reach.

http://www.slate.com/blogs/future_tense/2014/02/20/fcc_s_pla...

You need to qualify your statements with "in the US", because regulation has been effective elsewhere.

It's highly unlikely that the FCC will accomplish much in the way of enforcing net nutrality given that the agency has basically been captured by special interest.

Tom Wheeler, the new FCC Chairman, was a lobbyist and head of the NCTA for 5 years[1]. The idea that he will do anything to hurt the cable companies his old organization represents is ludicrous.

[1]http://www.nytimes.com/2013/05/02/business/tom-wheeler-telec...

> I disagree in terms of regulation: there is plenty of regulation on broadband internet, except it tends to discourage competition against broadband providers.

Well, yes - when I use the word "regulation", I'm implying the sort of regulation that we see with respect to Con Edison. They can't so much as change the way they itemize fees on my bill (!) without regulatory approval.

That certainly creates problems of another sort, but on the other hand, the actual service provided[0] and the pricing is pretty much a solved problem (as much as one can expect)[1].

I'm sure ConEd would like the sort of legislative protectionism that cable companies receive, with no requirements surrounding price or quality, but thankfully, they don't have that (and altering the status quo is always tough in general - in this case, inertia is on consumers' side).

Obviously regulation could be established to make things worse, but I think it's clear what kind of regulation we're looking for for the purposes of this discussion. And I'm intentionally deferring the question of how any of these policies would be enacted - I'm just comparing two alternatives states to the status quo.

[0] excluding customer service, which is going to be mediocre at best no matter what.

[1] Interestingly, pricing is partly "solved" due to the introduction of ESCOs, which are a way to privatize the supply chain (at the consumer's choice). "Free market" and "regulation" are not mutually exclusive - we could easily establish a regulated monopoly on last-mile copper wire while allowing multiple competing upstream providers.[2]

What leads me to believe we'll do better with new regulations? The belief, as chimeracoder suggests, that regulation can done right.

One point that caught my attention in an NPR report[1] on this dispute last week:

IHS's Broughton says there was a similar dispute in South Korea a few years ago, and it was ultimately arbitrated by the government.

I don't know the details of the South Korean dispute but I understand South Korea has pretty fast internet speeds.

http://www.npr.org/blogs/alltechconsidered/2014/02/20/280255...

South Korea also has highly invasive internet censorship.

Would you like that as well?

Oh of course not. We want the great parts but to leave the bad parts. Sometimes it's a package deal. I'll take what we've got, thank you very much.

South Korea's internet censorship has very little to do with their internet speeds. Saying "Well, if you want fast internet, it'll have to be censored" (as if our internet isn't fully monitored) is a false equivalence of the highest order.

Monitoring =/ censorship

If the gov't views the internet as it's possession, then it can make sense for it to be both fast and censored.

Well perhaps this may be the case in some scenarios, but that doesn't mean that the internet can't be censored when internet infrastructure is private, either. p = internet is public infrastructure q = censorship for some (but not all) p -> q for some (but not all) ¬p -> q q -> irrelevant

It's irrelevant to the question of fast internet.

To reiterate, you're saying that censorship and fast internet are a package deal, and if we want fast internet, we have to deal with censorship, amirite?

I'm saying that the political process which leads to things like gov't required fast internet may also lead to greater encroachment upon said internet.

It's an interesting way of thinking of things.

Yesterday I listened to a discussion comparing the less regulated and more stable banking system of Canada to the heavily regulated and crash prone US banking system. Everyone in the US says that they want Canada's banking system, it's a common refrain.

But do Americans want the political system which leads to such a banking system being implemented properly? As soon as you get to the Queen of England appointing US Senators, Americans tend to want to forget about the whole banking system thing.

I suppose the point is this: If you tell the US political system to implement Korea style fast and cheap internet, what you're going to end up with is something much different from what you wanted. The same political forces which got you to your current location will filter this legislation as well. And you may end up with something far worse than the status quo.

Canada's banks were still under the equivalent of Glass-Steagall restrictions at the time, and the government had kept them on a very short leash in many other respects. Your analysis is 180 degrees removed fom reality.

I don't know where the heck you're getting the nonsense about the Queen from either, the Canadian Prime Minister selects senators at his own discretion and the royal involvment in the process is just ceremonial.

Time for you to crack open our nation's history books!

Some things never change. I remember a article on South Korean broadband in Businessweek 15 years ago that ranted on about Seoul's paucity of elevators.

I'm guessing that package deal means we'll also be having to eat pickled cabbage and end our sentences with verbs if we want decent Internet speeds? I just hope we won't have to drink bubble tea, that stuff is gross.

Your whole post is based on a false premise, which is that customers have a problem with the status quo. I think internet companies have a problem with the status quo. They want somebody else to build the fat pipes crucial to their business. That's why Google fiber is a marketing and lobbying tool and not a real business.[1] But I don't think consumers here are any less happy than in other places. According to Akamai's state of the internet, the U.S. ranks #8 in average connection speed, ahead of places like Sweden. They have no reason to be. I've never heard someone outside my engineer friends complain about the lack of gigabit or even 50 Mbps+ connections. As long as its enough for Hulu, people are happy.

There are also other problems with your post. For example, we didn't sell off monopoly rights. Its been illegal for municipalities to grant exclusive franchises since 1992. Also, franchises don't come without strings attached. Here in Phila, Comcast has to run cable to a bunch of poor neighborhoods and charge $13 per month for service. Distribution requirements and regulated rates exist in the cable world.

[1] At the end if the day, if Google bought TWC instead, Google's average profit margin would go down.

Engineers are lead users for internet usage, they predict what others will want in several years. And even average users would become upset if they found out how much more they are paying than Swedes.

The U.S. beats Sweden in Akamai's study, so are we paying more to get more?

The US allegedly beats Sweden in average broadband speed which is a poor measure for most purposes. I don't think you will find any Swedes telling you that the US has better broadband.

From the FCC: "The United States is 26th out of 32 countries in the 15-25 Mbps speed tier (advertised) with an average stand-alone broadband plan price of $56.50. The lowest advertised price for stand-alone services is in Slovakia at $18, while the highest charges are found in Switzerland at $180"

Although they don't summarize the findings above 25 Mbps the story looks much worse.

Why do you rely on a study that uses advertised speed instead of actual speed like Akamai's? Akamai also has a category for 10 Mbps+ availability, and the U.S. ranks very well there too.

Because there aren't many international broadband studies that incorporate cost and I don't see any Mbps/cost breakdown in the Akamai study. The FCC report incorporates Ookla data (Speedtest.net) which uses actual broadband speed and costs submitted by users. It corroborates the conclusions of the FCC study.

With respect to availability: There is a world of difference between 1Gps@$1000/mth and 1Gps@$50/mth though both qualify as 1Gbps availability.

Finally >10Mbps is pretty broad bin today and fails to meaningfully differentiate "very" high speed services.

Us and Sweden are in the same ballpark, users everywhere are missing an order of magnitude or three. Gigabit is old and cheap, and after 10 gig last mile tech development stalled due to lack of interest. We should be running 10 gig on the fiber to apartments and 100 gig on the uplinks from buildings by now.

Instead, wired net connectivity has been overtaken by puny LTE...

"Your whole post is based on a false premise, which is that customers have a problem with the status quo. I think internet companies have a problem with the status quo. "

I recently had Google Fiber installed. Every single person I've talked to about it has responded with something similar to "Oh man, I wish I had that!" (That includes my tech-illiterate mom and the guy who came to install it who lives in a different city.) Of course consumers want faster, cheaper Internet connectivity.

"They want somebody else to build the fat pipes crucial to their business. That's why Google fiber is a marketing and lobbying tool and not a real business."

So, Google actually does the thing you said they don't want to do, and your response is, "well sure, but they don't really mean it?"

and the railroad network before that, etc

And don't forget: The telegraph industry! Neal Stephenson writes about it in "Mother Earth, Mother Board:" http://www.wired.com/wired/archive/4.12/ffglass_pr.html , which is fascinating and worth reading.

On one side there are irritated customers wondering why their service is so poor and why their bills are so high

One other "don't forget": a lot of people who pay attention are unhappy about franchise agreements and telco efforts to prohibit local utility fiber deployments by outlawing them at the state level (see https://news.ycombinator.com/item?id=7271044). Local municipalities that say, "Fuck it, we're going to roll our own" often find themselves forbidden because of successful lobbying at the state level.

>> this business model severely discourages new development (witness the "No one wants gigabit" rationale)

It's not that I want/need gigabit, it's that the only thing I can get in some locations is unusable. I'm looking at you, 1.5mb/768k DSL lines. It's barely enough for web surfing and useless for streaming video or doing things like gaming, video conferencing....

US achieved its greatness because of free flow of movement of people and their ideas. Roads and freeways and highways were plenty, fast and easily accessible, to both sellers of services/goods and consumers.

In this new world, internet access is the new road. Good internet access will generate new services/goods and encourage consumption by consumers. This will inevitably lead to snowball effect of creating more wealth for the society.

US shouldn't be held hostage by a few people in an executive suite in Philadelphia who want to make a few bucks so they can own a few extra mansions and a private jet or 2. US is a greater country than that.

One of the biggest missed opportunities of the 2009 government stimulus was a failure to heavily invest in new infrastructure that would pay dividends for the next 50-100 years. A government project to connect every household in America to gigabit Internet connection is more than doable, and would pay enormous benefits for individuals and corporations alike. It's not too late, although it would be much much harder today.

I agree, but the problem was that it would have run into a wall of local and state regulations, and the majority of such deployments would be getting fought over in court.

I thought the USA was the model of capitalism. However it appears the opposite, monopolies run rampant. The free market is killed rather than encouraged.

Monopoly is the apotheosis of capitalist desire, competition is merely the measuring stick.

Infrastructure is the quintessential example of a natural monopoly. Free markets encourage natural monopolies. A non-free market is required to avoid natural monopolies.

If you read the article it's plain to see that the monopolies are enforced and encouraged by the currently existing regulations. There is no free market in ISPs in the US.

You mean natural monopolies?

What are you talking about? Monopolies ARE capitalism run rampant.

Nonsense. Leftists conflate "state capitalism", cronyism or corporatism with real, free-market capitalism. Internet monopolies like Comcast or TWC enjoy are state-granted monopolies, not "natural monopolies" (which have never existed).

> Nonsense. Leftists conflate "state capitalism", cronyism or corporatism with real, free-market capitalism.

Leftists -- specifically, socialists -- created the name "capitalism" as a name for a real system which existed in the developed world in the 19th Century, in which the economic order was actively managed to favor the interest of the capital-holding class. The idea that this was system had something to do with "free markets" was part of the propaganda of defenders of that real, very much crony, system in response to that critique.

Its kind of weird to call the real thing which actually existed that "capitalism" was coined to refer to as anything other than real capitalism, while referring to amorphous, ill-defined, fantasy as "real capitalism".

"State capitalism" is a different thing entirely -- it usually shows up in critiques by leftists, particularly non-Leninist socialists, of Leninist(incl. Stalinist/Maoist)-style "Communism".

ITT: pedantry

Given that network effects and economies of scale are real things, on what grounds to you claim that natural monopolies don't exist?

The actual nonsense is the existance of the mythical " real, free-market capitalism", that nobody has ever seen in practice.

The "free market" that does exist is based on subsidies, bailouts, state protection, monopolies, a huge army, bullying, and resource grabbing (and past 300+ years of colonialism).

Heck, the "free market" is even more mythical than the laughable "really existing socialism".

Not to mention, taking an econ 101 supply and demand graph based on a quote 'widget factory' (actual textbook quote) and applying those ideas about marginal cost to utility economics where you build the grid once and then charge on subscription. Unironically. Earnestly, even.

The free market does not exist because the government stands in its way. To then go on to chide free market advocates for actions they do not support is disingenuous at best.

Businessmen are not interested in ideology, if they can use the government as a cudgel against their opponents, or a crutch to sustain them in hard times, they will.

Every time the government's power over the economy has been expanded, advocates of the free market have pointed this fact out and been ignored.

You can't break my spine and then blame me for being paralyzed.

No one broke your spine, it never existed and never will. It is imaginary.

Pretty sure my spine is real. Perhaps you meant to unpack that analogy first?

The "free market" is simply what exists absent interference. It ceases to exist when a single player or group of players asserts force over the others. The purpose of government is to reduce the effectiveness of violence as a strategy.

>The free market does not exist because the government stands in its way

So people should quit talking about it as it existed.

Not to mention that the same could (and WAS) said for communism ("it didn't exist because party bureucrats got in the way").

>You can't break my spine and then blame me for being paralyzed.

No, but I can point out that the unicorn pony you frequently talk about doesn't exist and never did in human history.

At various places and times throughout history, conditions of a largely free market have existed. It is not a "unicorn pony" (isn't that redundant and/or contradictory?). However, such conditions do not exist in most of the world now, and I don't see anywhere praxeologist or myself claimed otherwise.

> not "natural monopolies" (which have never existed).

Tell that to Standard Oil or The American Tobacco Company.

Do 10 minutes of research on Standard Oil. Not a great example.

I suggest spending more than 10 minutes of research on it. Standard Oil is quite literally the textbook example of a natural monopoly.

Based on my own non-expert personal musings on the subject, I would say that natural monopolies can and do exist in capitalism. It's just that under normal circumstances they would be difficult to sustain over a certain period of time based on several factors. Without some mechanic (often political) in place to prevent a new form of competition or technological progression from developing in the market, then a natural monopoly would likely be reduced by natural market forces.

If you look throughout just American history you can various examples of this. Big, strong companies that eventually fade away to disappear because of failure, lack of vision, or whatever other decisions that held the company back.

It's just that there is very few examples of a true free market, except for maybe on a local level, so that we'll likely never know for sure.

But of course no one advocates capitalism as an actual philosophical goal, they advocate that it's the best way to distribute our limited resources to get the most out of them. So if you accept the existence of natural monopolies it's hard to defend that the inefficiencies of regulating a monopoly to better capture externalities aren't less than the inefficiencies of personal profit maximizing behavior by the private entity with the monopoly.

Natural monopolies do exist, it's just that not all businesses claimed to be natural monopolies actually are.

If you can, imagine a watershed. The river, along with its tributaries, covers the most efficient downhill route from every point in the watershed to sea level. The natural monopoly on precipitation drainage is everywhere you see flowing water. Installing a second river does not work. Wherever you try to put it, the natural tendency is for the second river to drop right down into the original watershed. You end up with just one river again.

If you turn that idea into a graph, replace the ocean with the Internet, the power plant, the potable water and sewage treatment facilities, the highway system, or some other common resource, and replace raindrop targets with homes and businesses. There is only one optimal routing for roads, pipes, wires, and fibers. The surface path following that route is the natural monopoly. You can cram an awful lot of stuff on top of it if you wanted.

Once poles are up and tunnels in place, the hard part is over. If a second cable company runs a second cable, that's not a waste. The customer can now buy either of two cable service packages, or both. If there's already a hamster tube up on the poles connecting the house to the municipal rat supply, and someone wants to run another for squirrels, that's not a waste. People might not want their rats and their squirrels from the same tube.

So when people say cable and phone service are natural monopolies, they are wrong. It's the space those wires run through.

The analogy with a river doesn't work at all. Rivers don't act or engage in exchange. I see what you are saying on the most optimal path for water to flow, but that is then equivocating on "monopoly".

Monopoly is a single seller of a good or service. There is an economic definition that must be met for a "natural monopoly" to exist, which depends on the marginal cost of producing additional units and the fixed cost of maintaining the capital. Basically, just as you don't typically see two different rivers draining the same watershed, you don't see multiple firms serving the same natural monopoly market. That doesn't mean that it never happens; it just means that any challenger to the incumbent either displaces them entirely, acquiring their capital in the process, or goes out of business and yields its capital to the incumbent.

The lowest energy state of the natural monopoly is a single firm. Excited states can also exist, and are necessary for transitions to occur, but these states are less economically efficient. Customers see higher prices, and firms see lower profits. But given that a monopoly is in itself inefficient, it can still be worthwhile for a challenger to attempt to displace the incumbent, especially if they can operate at lower costs.

In a truly free market, the switch could be accomplished by an initial investment of sufficient size. In the context of the analogy, huge machines dredge a new channel for the river. For a time, water flows in both channels, but in the end, one dries up and you are left with one river again.

>I thought the USA was the model of capitalism.

You mean to tell me that the NYT has been lying to me all these years?

"AT&T is slowing Google Fiber deployment in Austin by denying access to its utility poles": This is why I am not getting Google Fiber yet?

Yup.

http://arstechnica.com/tech-policy/2013/12/why-att-says-it-c...

I have gigabit fiber internet and it doesn't feel all that much faster than my old Comcast 20Mbs connection because the bottleneck isn't my connection, it's the servers I'm asking for content from. For some things it's blazingly fast (downloading updates, torrenting, etc) but for general web surfing the benefit is barely noticeable.

I think that's why fiber adoption is slower than broadband adoption was. There's not much that you can do over a gigabit connection that you can't do over a 20Mbs connection for the average subscriber. Until some killer app that requires a gigabit connection to work starts to become popular, people just aren't going to demand it and adoption will continue to limp along.

A bit of a chicken and egg there though. Who's going to build services that use >20mbps in bandwidth if almost no one has that much?

I read some VC stories regarding Skype recently. At the time they were looking for series A (?), broadband penetration was still poor. Obviously they managed to get by until the broadband environment tipped to their advantage. The point is that you don't want to wait until it's obvious that gigabit is really coming. At that point you may already be way behind.

Agreed; the difference between downloading a movie in 10 minutes and 1 minute is rather miniscule.

Additionally, I don't think even streaming sites would generally require more than 50 mbps to watch comfortably. Not even talking about regular apps/static content.

As kyrra suggested elsewhere; you don't need the fat pipe for one stream, you need it for multiple streams. My household doesn't have cable TV so it's nice to be able to have the kids watching their stuff on Netflix in the other room while the wife and I are watching The Walking Dead in the living room. Without our FIOS that might not work out quite as nicely.

Does that killer app exist in Europe or Asia? Otherwise, why would they have a faster adoption rate than the US?

One of the arguments I saw for faster internet outside of a single-user app is when you have multiple people using a connection. If you have a family of 5 people all using a Cable connection in the US today, it is very easy to overload that line. And since upstream sucks on most home connections today (1.5 to 2 Mbps upstream), a single person uploading a video to youtube (or backing up data) will kill the connection for every other person using that line.

In the current ecosystem, I see gigabit connections being useful for multi-user connections where a lot of video is being streamed. Or using cloud storage solutions like Dropbox or Google Drive.

In the future, I could see other use cases. For what we know about Google's self driving cars, the first time a road is driven on (manual driving) it gathers a ton of data about that road. When you park in the garage, it may have to upload gigabytes worth of data for Google to process to make available to other cars. Doing that over the current connections most people have would be painful. If you have a large pipe for uploading lots of data, this becomes a lot more feasible.

For your first thought; that's one of the reasons as to why we love our FIOS connection, we can have multiple streams going at the same time and doesn't disrupt normal web surfing as well. It's a must with a large family.

As for the rest, I would say that it's hard for us to imagine the might single killer app with a gigabit connection because so few have it. Once a certain level of market penetration has happened, whatever that may be, there's no telling what things will come along once bandwidth is no longer a concern.

In brief, faster network connections are rolling out slowly because practically everyone that wants to do it faster is prevented from doing so thanks to statutory or logistical roadblocks thrown up by the incumbents, who have zero incentive to do it themselves because they enjoy a territory-designated monopoly.

The logical thing for most towns to do is to use their eminent domain powers to force the sale of easements and utility poles/tunnels to the municipality, and charge all utilities a flat rate based on pole capacity used. While two wires on a line of poles doesn't seem all that wasteful, two wires on two parallel lines of poles sounds pretty stupid. Digging up the street with two parallel trenches is even worse. As a consumer, I don't want a service provider in a position where it can keep its competitors from reaching by house via the best route possible. I want that access controlled by the cartel enforcer rather than any of the cartel members.

As a bonus, the town could allow residents or HOAs to hire qualified contractors to run and maintain their own cables to anywhere the poles/tunnels go. That way, if an ISP won't come to you, you can go to anything else you can reach.

thats a borderline moronic article.

starts by saying incumbents deny access even to city cabling and explaining the duopoly on most cities and ends by saying that allowing netflix to pass the fee to comcast subscribers is good because people will move companies and a free market is the answer.

is this april 1st or something? also extra moronic points for not understanding the first thing about isp to isp agreements and peer pricing.

The writer is suggesting that the "FCC can mandate right-of-way rules similar to those granted Google Fiber to all credible competitors".

He isn't saying that the "free market" will fix things by itself. He observes that there isn't a free market in providing internet access because the incumbents control things like telephone poles (to hang optical cable on) and municipalities are so addicted to the kickbacks they get from cable franchisees that they won't let anyone else in. The magic of the invisible hand only works if there is actual competition. Those supply and demand curve charts only describe reality if there aren't monopolists.

His claim is that if the FCC mandated right of way to anyone wanting to hook homes up then a company like Google would already be serving many more homes as incumbents wouldn't be able to block them as AT&T is doing in Austin.

This is true, but I'm not convinced that it is the right way to get proper competition here. Is the solution really to have 5 companies each run optical fiber down a street? Better would be for one company to run super-fast tubes, and then sell the bandwidth in bulk to ISPs. Maybe that is the business model that would emerge, but maybe not.

Leave it to the WSJ to blame a problem that needs old school trust-busting on excess regulation. They've got libertarian rabies.

Read the WSJ comments and you'll see another reason for the delay :-/

It is quite refreshing to see an article clearly stating the problem and the right solution. The barriers to competition needs to be broken, then the neutrality issue sorts itself out. We don't need to see entrenched cartels and monopolies 'fixed' through additional regulation imposing certain quality standards. That makes it hard for new players to enter the market and distinguish themselves, and more importantly it makes it hard for the dominant players to do the stupid things that will make their customers leave them faster for the competition.

In theory, if Google Fiber was widely available, would this drastically dilute the effects of net neutrality problems? Or are these two issues completely unrelated?

"Net-neutrality" means different thing to different people.

If by net-neutrality you mean the system where a paying broadband customer can consume any content, irrespective of that content owner ability to pony up to cash to broadband provider, then yes, these two issues are related. More competition in consumer broadband market would keep providers in check.

In an ideal world (at least my ideal), broadband providers would strive to do everything they can to deliver content to me. In current world, it's well documented that Comcast is passively allowing interconnects to fill up, causing congestion, and forcing content providers to pay up for better interconnect.

It becomes much harder to violate net neutrality in the face of competition. Here's an example:

Say you have two ISP options at your home. Firm A forces Netflix to pay $1/month for access to enough bandwidth to stream movies in HD to your house, because they want you to use janky-streams.com, which they own. Firm B keeps their peering infrastructure up-to-date without charging Netflix. Where does this leave you as a consumer?

All else equal, you'll switch over to Firm B. Why? Because Netflix works there even without paying a ransom for their data. It costs Netflix more money to provide their service to you on Firm A. They could pass that cost along to you directly (increase Netflix bill $1/month) or reduce the stream quality and blame the ISP. The end result is that you're paying more or getting worse service because of the Net Neutrality violation.

However, Firm A understands this logic. They know that charging Netflix would push consumers to Firm B. Thus they decide to play nice and pass along the data you already pay them for.

That's what I was suspecting, just wanted to be sure I understood things correctly.

'Your margin is my opportunity' -Jeff Bezos

I guess that only holds true when there aren't artificial barriers to entry into a market.

To complete the analogy, even more comically, to enforce the punishment by firm A against netflix, they probably had to make a capital investment in sabotage gear of perhaps 75 cents per month TCO.

Thats kind of how it was in telecom before they gave up on detailed long distance billing. You'd spend 4 cents/min on billing infrastructure which was OK when a LD call from CHC to NYC was $4/minute, not so great when Sams Club was selling LD cards for 5 cents/min.

Never forget that creating false scarcity of a non-scarce good takes a lot of time, effort, and money.

The problem is that instead, Firm B could just follow the lead and also charge $1/month for Netflix, leaving the consumer with no choice. Or they could charge $1/month for other streaming services and leave Netflix alone, so the consumer gets to choose between slow Netflix or slow Hulu & Amazon, with no chance to 'vote' with their wallets for fast everything.

This is the danger of oligopolies - even when they don't collude, their interests are in not competing 'too hard' to gain customers. If there are two firms, and each has roughly half of the customers, then the most money Firm A could get with no competition is 2x what they already have, so getting into a price war could easily result in lower total profits. For example, if they have a 50% profit margin, then lowering consumer prices by 25% (and thus cutting profits to only 25% margin) only breaks even if they are able to sign up every customer of their competition.

Free markets work best when there are many players and markets are easy to enter, as there is much more incentive to compete, which is not the structure of broadband markets.

If Google Fiber was widely available Google would be the one controlling traffic priority.

That is true, and why to the solution to the US's slow internet providers isn't "get Google fiber to everyone". It should be "make it easy for new companies to provide service, even if another fast service is already available".

Throw a couple of billions to create fibre last mile to every home in the USA that would be owned by the homeowners and it would go a long way to solving the problem. As a side effect it will also create that king of blue collar jobs that US needs right now.

I've got to imagine the availability of cheap and fast broadband influences property values in a given area. I've seen numbers like $5k thrown around; I'm tempted to think it's more like $10k-$30k.

When you multiply that amount by the number of homeowners in the country, that's a lot of potential influence. How could it be leveraged to improve broadband?

This is going to be a mess to slog through over the next 15 years until wireless speeds (and costs) leapfrog hardlines completely.

I've seen articles that it is just not feasible or even possibly to put all of the burden the wired networks carry onto the wireless ones. Besides the ugly data caps we have to deal with there, there's not enough room in the frequencies. Wireless is used as an excuse to not put in effort with wired and it's going to screw us.

There's plenty of room in the frequencies, even just using linear momentum encoding with directional transceivers. Add in MIMO/OAM and and suddenly there's even more bandwidth. the FCC / ownership of spectrum is a whole other chimera. Wireless today is like using a LED and a photodiode instead of a screen and a camera.

I think you mean OFDMA/MIMO, which is what LTE uses.

I should have Googled before I spoke, Orbital Angular Momentum, very interesting... http://lup.lub.lu.se/luur/download?func=downloadFile&recordO...

There's some debate about whether OAM demonstrations are just MIMO, and some suggestions that the eigenmodes are equivalent. Either way, there will be regulatory hurdles to more efficiently using the spectrum. I like the perspective put forward in this post: http://www.reed.com/blog-dpr/?page_id=117

I've been wondering for a while why this doesn't get brought up more whenever spectrum debates happen. I expect some amount of it chalked up to crappy fact reporting and lazy business attitudes, but you would think someone on CNN or the NYT could do 15 minutes of googling and actually bring some sense to the table...

come to Moldova, we have 100 mbit for 10$ month

Interesting...isn't boardband penetration somewhere around the 35% mark there?

no, every home has cheap and fast internet now

Reply on news.ycombinator.com