Population Density

I think it's a dodge to answer the question about why U.S. broadband is so much worse than Japanese or South Korean broadband by referring to America's lower population density. If the issue were that we have excellent internet service in some places, but it's super-slow in Wyoming then, sure, you'd say it's the density. But the population density of Southern California or, say, New Jersey is pretty high and you don't get Asian-style broadband there, either. For that matter, you don't get Asian-style broadband in Manhattan.

The relevant issues here are regulatory in nature -- Japan has smart regulations that produce quality service, whereas the U.S. has regulations that are good for incumbent telecom firms. It's true that there would still be a real residual issue related to people who live in very low-density areas, but that's a separate issue from why the majority of Americans who live in metro areas can't get decent broadband.

Comments

A link to an analysis of Japan's "smart regulations that produce quality service" would be nice.

Posted by ostap | March 21, 2008 12:05 PM

The people who blame our low population density also need to account for the superiority of Canadian broadband in comparison to ours.

Posted by Phaedrus | March 21, 2008 12:11 PM

Holy cow!

First the correction on how the sixe of the House is set, and now a response to fallacious arguments about telecom policy. Matt has taken up reading some of the comments.

What's next, Matt? Correct spelling?

Posted by Brian | March 21, 2008 12:12 PM

Holy cow!

First the correction on how the size of the House is set, and now a response to fallacious arguments about telecom policy. Matt has taken up reading some of the comments.

What's next, Matt? Correct spelling?

Posted by Brian | March 21, 2008 12:13 PM

Actually the eu telco regulations have been successful by mimicking the US, not the other way around. The only complaint would be that state regulators have set the wholesale rates high enough to stifle competition among last-mile competitors.

Posted by Nicholas Beaudrot | March 21, 2008 12:15 PM

"Correct spelling?" - In your dreams, Brian.

Posted by fostert | March 21, 2008 12:16 PM

It's interesting that the US has a slightly overrall population density than Estonia.

Posted by nu | March 21, 2008 12:30 PM

Way back at the 2000 Builder conference, Alan K'necht predicted that the U.S. would soon face a significant technology disadvantage with the developing world. The U.S. telecom infrastructure is still largely built from copper wire, and upgrades need to be integrated into that system, whereas in the developing world, fiber optic and wireless telecom networks are being built from the ground up.

Posted by Toady | March 21, 2008 12:31 PM

higher. slightly higher.

Posted by nu | March 21, 2008 12:32 PM

Knapp's second reason is specious, too -- i.e. multiplicity of regulators. A backbone provider's line might stretch 20 Km, but isn't subject to local jurisdiction except as a matter of physical infrastructure. Last-mile providers are the only ones, AFAIK, who have to deal with municipalities. And those last-mile providers are the ones who asked Congress to grant them nationwide franchising rights, and, failing that, have been pushing states to grant statewide rights, thus cutting municipalities out of the picture. In other words, both of Knapp's points are telco talking points.

Posted by Rick Karr | March 21, 2008 12:40 PM

So... let's see: South Korea has great broadband access because they're a fairly dense country. Canada, by contrast, has great broadband access because the policymakers figures that in a large country with a dispersed population, it was a good idea to have fast communication technology available everywhere to help bind the country together.

I think we can see where this is going, when it comes to explaining why the US doesn't have broadband of the same caliber, even in its dense regions. It's because policymakers don't care enough to make it a priority.

Posted by Tyro | March 21, 2008 12:42 PM

I'm not familiar with Japan's telecom regulations, but I find it hard to believe that a regulatory system in Japan exists that doesn't cater to the bottom lines of corporations. Either the corporations know enough to provide decent service AND pad their bottom line or nobody's noticed the telecom's are dangerously overstretched and they'll collapse in a glorious display of modern-day harikiri.

Posted by Jeremy | March 21, 2008 12:47 PM

My repairman (ATT, LA) tells me my neighborhood will soon be getting fiber optic cable. Then my internets will zoom, yes?

Posted by bobbo | March 21, 2008 12:52 PM

I think it's a dodge to answer the question about why U.S. broadband is so much worse than Japanese or South Korean broadband by referring to America's lower population density.

It's probably a combination of factors, and population density is one of them.

For that matter, you don't get Asian-style broadband in Manhattan.

Irrelevant. Even if certain areas served by providers have a high population density (e.g., Manhattan) other areas may not. It's the overall density of the total population the provider serves that matters most, not the density of particular areas.

The people who blame our low population density also need to account for the superiority of Canadian broadband in comparison to ours.

Canada's population is concentrated in and around a relatively small number of urban areas. The vast bulk of the country's geographic area has virtually no people at all. That is not the case in the United States.

Posted by Mixner | March 21, 2008 12:55 PM

Out of curiousity, how does broadband available in Canada differ from that available in the States? I was under the impression that the service and rates available in Japan were vastly superior to both.

Posted by Paper Mac | March 21, 2008 12:57 PM

Japan has smart regulations that produce quality service, whereas the U.S. has regulations that are good for incumbent telecom firms

I'd love a post explaining what is the difference between the smart regulations there and the bad ones here.

Posted by Al | March 21, 2008 12:58 PM

"It's because policymakers don't care enough to make it a priority."

Or that there is little demand for faster broadband in the US, so broadband providers don't see the point of investment. The policymakers don't care because their constituents don't care. I purchased to lowest level of broadband available from my cable provider and it's plenty fast enough for my needs.

Posted by JordanT | March 21, 2008 12:58 PM

Depends on how you define "constituents". If people didn't care about broadband at all, they wouldn't have traded up from dial-up in the first place, so I think Jordan's personal example is a bit irrelevant.

I'd say the answer is pretty clear, though: Internet consumers have far, far less sway with Congress than the telecoms do. Unlike Japan--where NTT was pretty much at the government's mercy--the oligopolistic telecoms can get away with doing pretty much whatever they want, as long as the public doesn't get too ticked about it.

(Even if broadband were a voting issue, consumers tend not to understand why computer speed would be bottlenecked in the first place. They'd probably just blame it on the computer itself. Most of the applications they use the speed for are of dubious legality anyway, so how could they complain?)

Al: the smart regulations, from what I understand, were about unbundling DSL from local telephone service and creating competition between DSL providers. Telecoms are natural local monopolies. At best, they're oligopolies. Unless you foster competition, they're going to exploit those monopolies to deliver poor service at exorbitant prices.

Fortunately, that's exactly what Japan did, and they followed it up by dumping a ton of money on anybody who took that critical step of ripping out the DSL and putting fibre in all the way to the home. Considering the big bottleneck has always been that "final mile", that was a critically important and vital move.

(And that's the problem in the U.S: those near-unchallenged local monopolies of companies that don't want pay to install fibre to the home, thank-you-very-much.)

Posted by Demosthenes | March 21, 2008 1:23 PM

Mixner, your two points about Manhattan and Canada contradict each other entirely. When discussing manhattan, you claim that local occurences of density don't matter. Then discussing Canada, you claim that broadband has been rolled out to such a huge country because it has local density.

Much like your "contributions" to discussions about transit policy, you simply can't think of things in quantitative terms.

The policymakers don't care....

Yes, that is precisely the problem. We can come up with reasons WHY something is (policy makers make plenty of policy decisions that constituents don't care enough to support or oppose, so the rest of your point is orthogonal to the matter), but the point is that the country lacks the well that Canada and South Korea had to push greater broadband speed through.

Posted by Tyro | March 21, 2008 1:26 PM

Depends on how you define "constituents". If people didn't care about broadband at all, they wouldn't have traded up from dial-up in the first place, so I think Jordan's personal example is a bit irrelevant.

I'd say the answer is pretty clear, though: Internet consumers have far, far less sway with Congress than the telecoms do. Unlike Japan--where NTT was pretty much at the government's mercy--the oligopolistic telecoms can get away with doing pretty much whatever they want, as long as the public doesn't get too ticked about it.

(Even if broadband were a voting issue, consumers tend not to understand why computer speed would be bottlenecked in the first place. They'd probably just blame it on the computer itself. Most of the applications they use the speed for are of dubious legality anyway, so how could they complain?)

Al: the smart regulations, from what I understand, were about unbundling DSL from local telephone service and creating competition between DSL providers. Telecoms are natural local monopolies. At best, they're oligopolies. Unless you foster competition, they're going to exploit those monopolies to deliver poor service at exorbitant prices.

Fortunately, that's exactly what Japan did, and they followed it up by dumping a ton of money on anybody who took that critical step of ripping out the DSL and putting fibre in all the way to the home. Considering the big bottleneck has always been that "final mile", that was a critically important and vital move.

(And that's the problem in the U.S: those near-unchallenged local monopolies of companies that don't want pay to install fibre to the home, thank-you-very-much.)

Posted by Demosthenes | March 21, 2008 1:31 PM

Depends on how you define "constituents". If people didn't care about broadband at all, they wouldn't have traded up from dial-up in the first place, so I think Jordan's personal example is a bit irrelevant.

I'd say the answer is pretty clear, though: Internet consumers have far, far less sway with Congress than the telecoms do. Unlike Japan--where NTT was pretty much at the government's mercy--the oligopolistic telecoms can get away with doing pretty much whatever they want, as long as the public doesn't get too ticked about it.

(Even if broadband were a voting issue, consumers tend not to understand why computer speed would be bottlenecked in the first place. They'd probably just blame it on the computer itself. Most of the applications they use the speed for are of dubious legality anyway, so how could they complain?)

Al: the smart regulations, from what I understand, were about unbundling DSL from local telephone service and creating competition between DSL providers. Telecoms are natural local monopolies. At best, they're oligopolies. Unless you foster competition, they're going to exploit those monopolies to deliver poor service at exorbitant prices.

Fortunately, that's exactly what Japan did, and they followed it up by dumping a ton of money on anybody who took that critical step of ripping out the DSL and putting fibre in all the way to the home. Considering the big bottleneck has always been that "final mile", that was a critically important and vital move.

(And that's the problem in the U.S: those near-unchallenged local monopolies of companies that don't want pay to install fibre to the home, thank-you-very-much.)

Posted by Demosthenes | March 21, 2008 1:34 PM

I agree with several commenters above. Matt, here is the FCC's most recent unbundling order. It's about 180 pages long. Here is the next most recent unbundling order -- beware, that's a 576-page PDF file.

So in what specific way do the FCC's rules differ from Japan's?

Posted by Stuart Buck | March 21, 2008 1:40 PM

I don't buy the density excuse. I was just visiting my sister. She lives outside of Terlingua, TX, which has to be one of the least dense areas in the U.S. Her DSL is easily available and just as fast as mine in Philly.

Posted by gbh | March 21, 2008 1:41 PM

I agree with several commenters above. Matt, here is the FCC's most recent unbundling order. It's about 180 pages long. In what specific way do the FCC's rules differ from Japan's?

Posted by Stuart Buck | March 21, 2008 1:42 PM

I dunno, my broadband in Ann Arbor, MI is about as good as what I got in Osaka, Japan a couple years ago. I didn't have FTTH in Japan, though - that stuff is awesome.

Posted by Mr. Noah | March 21, 2008 1:42 PM

"If people didn't care about broadband at all, they wouldn't have traded up from dial-up in the first place, so I think Jordan's personal example is a bit irrelevant."

That's because dial-up users at the time were hoping for something better to come along. Some even went though the enormous expense to have an ISDN or T1 line installed. Even with much simpler web pages back then, using a dial-up connection sucked. Busy signals trying to dial in, the need for local numbers, tying up a phone line, random disconnects, high pings for online gaming and painfully slow speed download speeds.

When there was a switch from 28K modems to 56K modems the difference was pretty small. It wasn't until the comparably huge leap of broadband internet did it stop being frustrating to do anything online. My point is a 2X increase in broadband speed would likely do 0 for the average consumer. That's why most just elect for the cheapest service available and don't pay extra for a better connection.

Posted by JordanT | March 21, 2008 1:44 PM

There might be regulatory differences between the US and Japan, of course, although I'm very skeptical that any generalist journalist is going to have any idea how to figure out those differences. In any event, you also have to take into account the fact that loops in Japan and Europe are apparently much shorter than in the United States. This is crucial, because longer loops means that DSL just won't work as well -- transmission degrades as the loop gets longer. See, e.g., this: http://www.cedmagazine.com/technology-globetrotter.aspx

CED: Do international telcos have the loop length issues you've mentioned to me as an issue for state-side telcos?

Werner: The telephone plants in most of the world are better than they are here. They're shorter. Japan has very short loop lengths. A thousand feet, give or take. They didn't use twisted pair there. They used a copper wire that's flat. So it doesn't have quite as good propagation for DSL as in the U.S., but still, they have a better DSL platform in Japan. They're offering higher speed services over copper in Japan, but they do have some reach issues. Not all customers can get those speeds. But definitely they have an advantage over U.S. DSL providers.

Europe is the same. Loop lengths are shorter—a half to a third the length of U.S. loops.

CED: So 3,000 or so feet?

Werner: Suffice it to say that they don't have any of the 12,000- to 18,000-feet loops like we have here. In general, the loops are shorter, the plant quality is better, and as a result, they can offer higher speeds.

Posted by Stuart Buck | March 21, 2008 1:56 PM

Mixner, your two points about Manhattan and Canada contradict each other entirely. When discussing manhattan, you claim that local occurences of density don't matter. Then discussing Canada, you claim that broadband has been rolled out to such a huge country because it has local density.

You are confused, as usual. As I said, the most important density measure is the density of the overall population served by the provider. High-density Manhattanites subsidize the costs of providing service to lower-density areas served by the same providers. With respect to a comparison between the U.S. and Canada, the Canadian population is much more densely distributed than the U.S. population, and thus cheaper to serve. 90% of Canadians live in a narrow strip of land within 100 miles of the U.S. border. The vast bulk of the country is an icy wasteland with virtually no inhabitants.

Posted by Mixner | March 21, 2008 1:57 PM

I agree with several commenters above. Matt, here is the FCC's most recent unbundling order. It's about 180 pages long. In what specific way do the FCC's rules differ from Japan's?

Dear Stuart,

You have no fricking idea what you're talking about. There is no unbundling for Internet service in the United States. I don't have time to post links right now, but can do so later if you insist. Trust me on this.

Posted by J.B. | March 21, 2008 1:58 PM

I haven't been to Asia but I travel in Europe a lot, and my sense is that the rest of the advanced world is developing a better infrastructure than us in just about every area. And at some point this might become a concern to us (LOL).

Like you said, regulation in the rest of the world mandates what the country is going to need, while regulation in the U.S. protects the existing players and actively discourages innovation.

One good example is the prevalence of "smart cards" in Europe It costs money for merchants to upgrade, there are lots of merchants, so the U.S. won't tell them to spend money. Many might want to, since it'd be more efficient and profitable for them in the long run. But no one can do it unless everyone does it.

In Europe, the government simply makes everyone do it, and then everyone is glad they did.

Health care is the same way, and I wonder if our political culture (neither party thinks outside this particular box) is dooming us to obsolesence.

Posted by Alan in SF | March 21, 2008 1:59 PM

Alan, part of it is the higher costs of labor in Europe. Because hiring someone is so expensive, there's a greater incentive to adopt technologies, like smart cards, that make automation easier.

Mixner, enough of the US is very dense that the comparison to Canada is a good one. While you say that 90% of canada exists over a "narrow strip", the truth is that this strip is very very long. Canada prioritized fast broadband even out to far-flung places like New Brunswick precisely because their population is spread out-- in the same way that the united states has population centers which are widely separated from each other.

Your point about having to subsidize smaller markets is irrelevant, presumably dragging down the larger ones, is irrelevant, here, particuarly given the fact that broadband was selectively distributed in the first place... we would equally expect canadian or korean-level broadband services in precisely those very dense areas of the USA, in the same way that broadband itself was available in denser areas before it arrived elsewhere. But, once again, thanks for your "contributions" about things you clearly have trouble puzzling out logically. It's like you throw an irrational, unthinking temper tantrum whenever population distribution issues are discussed.

Posted by Tyro | March 21, 2008 2:12 PM

J.B., I've practiced in this area of law for several years, and published scholarly articles. You said there is no loop unbundling in the United States, which is false. Maybe you meant to say there's no line sharing, which refers to the requirement that ILECs unbundle the HFPL (high frequency portion of the loop, over which DSL transmission occurs).

If that was what you had said, you'd be right. Line sharing was required by the FCC's Line Sharing Order in 1999, but that order was vacated by the D.C. Circuit's USTA I decision in 2002, and the FCC didn't try to re-impose the rule in the Triennial Review Order. All that means, of course, is that competitive DSL providers need to lease the entire loop on an unbundled basis, which they can still do under Rule 51.319.

Do you have any evidence that broadband deployment in the US was on a different trajectory between 1999-2002?

Posted by Stuart Buck | March 21, 2008 2:27 PM

Tyro,

Mixner, enough of the US is very dense that the comparison to Canada is a good one.

Please substantiate this claim with actual population density data.

While you say that 90% of canada exists over a "narrow strip", the truth is that this strip is very very long.

But it's very, very narrow. The vast majority of Canadians live in a narrow strip of land that comprises only a small fraction of the total area of Canada. Americans, in contrast, are much more widely distributed across the geographic area of the United States.

Your point about having to subsidize smaller markets is irrelevant, presumably dragging down the larger ones, is irrelevant, here,

No it isn't. It's relevant for precisely the reasons I explained.

... particuarly given the fact that broadband was selectively distributed in the first place... we would equally expect canadian or korean-level broadband services in precisely those very dense areas of the USA, in the same way that broadband itself was available in denser areas before it arrived elsewhere.

Huh? Why would you expect that? Manhattan's cable provider, for example, Time Warner, provides service in 27 states and many areas of the country with population densities very much lower than Manhattan. Manhattan subscribers subsidize the costs of serving those low-density areas.

But, once again, thanks for your "contributions" about things you clearly have trouble puzzling out logically. It's like you throw an irrational, unthinking temper tantrum whenever population distribution issues are discussed.

Thanks again for yet another ignorant, irrational, moronic Tyro post.

Posted by Mixner | March 21, 2008 2:31 PM

"Why would you expect that? Manhattan's cable provider, for example, Time Warner, provides service in 27 states and many areas of the country with population densities very much lower than Manhattan. Manhattan subscribers subsidize the costs of serving those low-density areas."

We would expect that because it's what happened with the first stages of DSL. Or what's happening with, say, Verizon's fiber.

And even if it's true that it's unprofitable to provide high-speed in even dense areas, because of something to do with the structural composition of US firms then...That's still a regulatory issue.

Posted by jack lecou | March 21, 2008 2:52 PM

Mixner, look at a fucking map, will you? Draw a line from Vancouver to St. Johns, passing through all the major cities. Then look at California, which has the same population. Then just let it settle for a while, maybe with a nice cup of tea.

Posted by nolaboyd | March 21, 2008 3:37 PM

I was in Strasbourg last month. The stops on the new tramlines had adverts for 100 mb/sec fibre-optic broadband. Wake up: what Paul Krugman wrote on the growing US lag is true.

Posted by James Wimberley | March 21, 2008 4:33 PM

All that means, of course, is that competitive DSL providers need to lease the entire loop on an unbundled basis, which they can still do under Rule 51.319.

So... in other words, there is no unbundling of Internet service.

If I want to get residential DSL in the upscale, densely populated urban area where I live, I can get Verizon or... Verizon. In answer to your initial question, this is not the case in Tokyo.

Posted by J.B. | March 21, 2008 4:41 PM

nolaboyd, no you look at a fucking map. What about California? Do you have a hint of a clue of what point you think you're trying to make?

Posted by Mixner | March 21, 2008 4:46 PM

P.S. To satisfy your lawyerliness I thought I'd throw this in: See also Appropriate Framework for Broadband Access to the Internet over Wireline Facilities, Report and Order and Notice of Proposed Rulemaking, 20 FCC Rcd 14853 (2005).

Posted by J.B. | March 21, 2008 4:51 PM

"nolaboyd, no you look at a fucking map. What about California? Do you have a hint of a clue of what point you think you're trying to make?"

Uh... I understood nolaboyd's point* just fine. Maybe you could elaborate where you're disagreeing with it?

(*Namely-
Canada: ~32M people concentrated in large cities along a narrow-but very, very long-band;
California: ~36M people, concentrated in large cities along a similarly narrow-but much shorter-band. Get it?)

Posted by jack lecou | March 21, 2008 5:18 PM

So I take it you don't have any evidence as to the effect of US line sharing rules (which you claim make the difference in Japan)?

It's very difficult to support the claim that something like broadband deployment is entirely due to regulatory differences. There's a lot more going on here than just regulation (such as the average loop length).

Posted by Stuart Buck | March 21, 2008 5:25 PM

Stuart-

I read the interview you linked to, and didn't really find the comment about loop length to be all that interesting. It's possible long US loop length may have an impact on top speed in some areas, but that doesn't really address the issue of competition-- which seemed to be the much more interesting part of that interview. In Japan, competition is fierce, with 10 or 20 providers fighting over the same area (presumably thanks to aggressive unbundling--ISPs there can rent a copper pair from the telco for, what,

In any case, average loop length may be the same geographical red herring by another name. Loops in Eastern Oregon might all be 18,000 feet, sure, but what's the typical loop length in Manhattan or LA?

And, to the extent that even dense areas suffer from loop length problems, it's still hard to see how long loops affect fiber much. Possibly loop length has some effect on initial rollout cost, but then, if long loops are the problem, shouldn't we be working on policies to encourage the installation of more local switching centers or something?

It's policy--and competition--not geography or destiny.

Posted by jack lecou | March 21, 2008 5:53 PM

(*Namely-
Canada: ~32M people concentrated in large cities along a narrow-but very, very long-band;
California: ~36M people, concentrated in large cities along a similarly narrow-but much shorter-band. Get it?)

No, I don't get it. What is your point? Hint: Canada is a country; California is not.

Posted by Mixner | March 21, 2008 5:55 PM

Oops- end of first paragraph got cut off:

...In Japan, competition is fierce, with 10 or 20 providers fighting over the same area (presumably thanks to aggressive unbundling--ISPs there can rent a copper pair from the telco for, what, less than $2.00? You seriously don't think that makes a difference?). In Puerto Rico, not so much. (128K for $20?)

Posted by jack lecou | March 21, 2008 5:56 PM

Canada is a country; California is not.

I see... So, if California were to secede from the Union, Verizon would start offering decent high speed internet there?

Posted by jack lecou | March 21, 2008 6:00 PM

Mixner-

I don't mean to be snarky. The point is that it's about policy, not average density. (A point you now seem to be implicitly conceding when you say it has something to do with administrative boundaries, not population distribution.)

If you want to argue that circuitiously subsidizing low-density areas by withholding innovation and competition from urban users is good policy, go for it. Just don't pretend it's some kind of geographic imperative.

Posted by jack lecou | March 21, 2008 8:16 PM

"It's the overall density of the total population the provider serves that matters most, not the density of particular areas."

This doesn't even make any sense. We're talking about laying down actual physical cable - not dangling it in the air over over thousands of square miles. There is absolutely nothing stopping Time Warner for laying cable in Manhattan to provide Manhattan with Asian speed while still ignoring the rest of their territory.

Not to mention that the additional income they would make from doing that could cover the cost of providing higher, if not highest, speeds in the rest of the territory.

Unless of course there are regulatory concerns about providing the exact same class of service to everybody regardless of the technological difficulty of doing so. That would be a legitimate complaint.

Provide a cite for that or STFU.

Mixner is a fucking troll. He comes in, picks some random element of a discussion, then nit picks and semantically twists it to death until people get tired of talking to him.

Don't waste your time.

Posted by Richard Steven Hack | March 21, 2008 8:50 PM

Not to mention that the additional income they would make from doing that could cover the cost of providing higher, if not highest, speeds in the rest of the territory.

The problem is that letting monopoly cablecos and telcos provide the service is that we end up with 1-6Mbps lines for $60. And we stay there.

There's no reason that providing cheap 100Mbps to Manhattan would be more profitable for the monopolies. They'd either have to keep their pricing structure and charge north of $300/mo for it; or they'd have to go to all the expense of upgrading their equipment, and still only be getting $60 or so a month, from the same subscribers who seemed happy paying the same price before the upgrade. There's just not much incentive for lowering costs or improving quality.

The way toward Japan-style, $30/mo 100Mbps involves making things far less profitable for the monopoly owners of the copper: you need to unbundle it, and let independent ISPs lease residential pairs and telco rack space for very low, regulated prices.

All of which probably does leave less cash in the telcos hands for subsidizing rural service.

But, I think it's not only pretty debateable whether they do this much anyway, it's extremely debateable whether forcing urban economic zones to remain in the internet dark ages is a very sensible way to subsidize anything in the first place.

Posted by jack lecou | March 21, 2008 11:02 PM

So I take it you don't have any evidence as to the effect of US line sharing rules (which you claim make the difference in Japan)?

First off, nice of you to concede that the US doesn't have unbundled Internet access. As to this question, we're dealing with social policy here, not physics. Exact proof is never available. But is seems pretty obvious that competitive environment is going to lead to better prices and services than a noncompetitive environment.

There is also evidence to support this fairly obvious point. Certainly the ITU has seen unbundling the local loop as a key to effective broadband competition, stating that "[m]any successful broadband economies have achieved their high penetration levels precisely because of local-loop unbundling and strong competition." (p. 28) Additionally, the ITU's country-specific studies of Korea (p. 12) and Japan (p. 28) both cite open access as key drivers of broadband competition. ITU studies of the successful broadband strategies of Canada, Iceland, and Hong Kong have also noted their open access policies. Meanwhile the US has languished with broadband monopolies or duopolies in most local markets, sliding further behind in the rankings every year. I'd say the evidence is pretty damned strong.

Posted by J.B. | March 22, 2008 1:39 AM

Look, you want a simple, if incomplete explanation?

WWII.

Many of the countries we're talking about either had no phone service to speak of until lately, or got bombed back into the stone age about 60 years ago. Here, we've got lines in use that were personally laid by Alexander Graham Bell!

It's easy to do things right, if you're doing them for the first time, or redoing them.

I'm not saying this is the total explanation, but on a social/cultural level, too, not in living memory having had a war press society's 'reset' button makes a big difference. Institutional inertia has put down huge roots in this country, while other countries got those roots plowed up by carpet bombing.

Try to come up with a way to press that reset button without killing a heck of a lot of people. It would solve an amazing number of problems.

Posted by Brett Bellmore | March 22, 2008 8:18 AM

Many of the countries we're talking about either had no phone service to speak of until lately, or got bombed back into the stone age about 60 years ago.

Of course this is total nonsense: have a look at this map of UK broadband coverage in 2002 and this map of of UK broadband coverage in 2005. This was achieved within a time-frame of three years, it all runs on copper lines dating back to Victorian times and you get speeds between 2-24 Mbps for £7.50 to £20 a month with a choice among 50 providers.

I always thought these conservative/libertarian types had a "can do" attitude when it comes to innovation, but it seems they would rather blame god, the weather or their neighbour's cat than admit that other countries might have on occasion gotten something right, while the US has messed up.

But there's hope: fiber-optic lines are being rolled out as we speak and it's a whole new ball-game - that is, for those interested in change for the better rather than finding excuses for slow speeds and high prices.

Posted by novakant | March 22, 2008 9:35 AM

Many of the countries we're talking about either had no phone service to speak of until lately, or got bombed back into the stone age about 60 years ago. Here, we've got lines in use that were personally laid by Alexander Graham Bell!

Nonsense. As per the ITU studies I posted above, Korea and Japan both lagged way behind the U.S. in broadband and Internet connectivity as late as 2000. Both then adopted sensible regulatory polices and went roaring to the front of the pack. As I said in my initial post: this is not that complicated.

Posted by J.B. | March 22, 2008 9:55 AM

First off, nice of you to concede that the US doesn't have unbundled Internet access.

No concession was necessary, as I never said otherwise. I said that there is "loop unbundling" in the United States, which is absolutely true. Again, in the United States, "loop unbundling" means being able to lease the entire local loop; "line sharing" means being able to lease just the HFPL while the ILEC continues to provide POTS.

You cite an ITU report that supposedly "cites open access as key drivers [sic] of broadband competition." The opposite is the case. Page 64 says that "[l]ocal loop unbundling, although legal, has not been widely implemented due to commercial barriers." Then page 67 states:

Although local loop unbundling is possible in Korea, most broadband providers are facilities-based. . . . The lesson here is facilities-based competition, one that even developed countries can learn from. Too many
nations have focused on local loop unbundling as the way towards creating a competitive broadband market. The results have not generally been very successful. Korea has shown that facilities-based competition, even in a mature network, works.

Indeed, given that people seem to use the term "loop unbundling" to mean different things, it's very unclear to me whether South Korea and Japan's regulatory approach involves actual line sharing (what you call "unbundled Internet access") or whether those reports really mean it when they say "loop unbundling." Even as a telecom expert, I wouldn't dare to assess South Korean or Japanese telecom regulation without being able to read the relevant languages. But if Japan and South Korea really do have just "loop unbundling," then regulatory differences can't be as important as you think, because the United States has loop unbundling too. (Again, see rule 51.319.)

Posted by Stuart Buck | March 22, 2008 11:45 AM

Those quotes showing the opposite of what "J.B." claimed were from the South Korean report.

Posted by Stuart Buck | March 22, 2008 11:49 AM

Those quotes showing the opposite of what "J.B." claimed were from the South Korean report.

Posted by Stuart Buck | March 22, 2008 11:50 AM

Re: That's why most just elect for the cheapest service available and don't pay extra for a better connection.

I think you just hit the nail on the head. We don't have better broadband because there isn't sufficient demand for it. Sure, there's some technogeeks here whining the they want it, but there's just not enough such people.

Re: Health care is the same way, and I wonder if our political culture (neither party thinks outside this particular box) is dooming us to obsolesence.

American healthcare is in no danger of becoming technogically obsolencent: it's as cutting edge as can be. The problem is the equity of its distribution.

Posted by JonF | March 22, 2008 11:42 PM

We don't have better broadband because there isn't sufficient demand for it.

Bull. It's the nature of telecom that technology must lead demand. Otherwise we'd still be on rotary-dial POTS. Or telegrams.

The problem here is not that consumers choose a cheap-ish level of service. The problem is that "basic" in Nagasaki means something like 6Mbps at $15, while "basic" in LA means 256kbps for $30.

Stuart-
Agreed that more facilities-based competition would be excellent. Too bad about the 700Mhz auction...

Posted by jack lecou | March 23, 2008 7:34 AM