Dark Fiber: A Case In Point

Anonymous Coward writes "CNN has posted a story regarding the overabundance of fiber lines that were laid during the 90s gold rush along Oregons Interstate 5 corridor. While over 140,000 miles of fiber has been laid 95 percent of the fiber goes unused and roughly half of the companies who laid the fiber are now gone. The article goes on to further say that even with all that fiber, there is little availability to the consumer because either the local connections aren't there or, because of monopolization by phone companies, too expensive. Even for businesses."

Comments

Old News

[geekee]

This is old news. Companies laid a lot of fiber at once knowing it wouldn't be used immediately. Given the cost to lay the fiber relative to the cost of the fiber itself, this is not unreasonable. The fiber is not lit currently because the tranceivers are very expensive.

Yes way

[siskbc]

Might want to grab an econ book - there are many industries in which a dramatic increase in quantity of product would drop prices so much that the overall net revenue would be lower at higher quantity.

Examples include farming (hence we actually pay farmers to grow nothing), steel (at least now), oil (otherwise OPEC wouldn't set production quotas), and, yes, bandwidth.

To follow your argument, why then AREN'T the phone companies selling the extra bandwidth? It isn't the demand - I would like some cheap bandwidth. It isn't the lack of fiber - as the article says, there's a lot unused. It wouldn't be that hard to tap, especially since most consumers would be willing to pay for reasonable install costs.

No, the reason is the phenomenal price drop that an increase in quantity would bring, nothing more. And you're right, it's not about "evil" phone companies - it makes good sense to do what they're doing. I've never known a company want to DROP their prices, certainly.

Re:Yes way

[milo_Gwalthny]

Hooold on a second. Don't throw out that "grab an econ book" canard without saying something that would actually be in an econ book.

There are two things going on here:

1) The telephone companies have dark fiber. You are saying they are not lighting it because that would lower their pricing power. How is that? If you mean they aren't going to double the bandwidth to a particular customer without increasing the price, I agree, but I don't think that's the issue--if they don't have pricing power it's because they're regulated, but it sounds like the Oregon PUC might allow them to make money on increased bandwidth. I think the increase in bandwidth we're talking about here would be to *new* customers. A new customer to the telephone monopoly doesn't change their pricing power at all. They may have to lower their prices to get new customers, but that's a different issue--it's not the increase in quantity that would lower prices, it's the lower prices that would be needed to bring about an increase in paying customers. This is a problem that every high fixed-cost, low variable-cost business has (ie. airlines, telcos, pharma companies, software companies, etc.): how do we get the people willing to pay a lot to pay a lot (and cover the fixed expense) while allowing the people who only want to pay a little to pay a little (and cover a little more than the marginal cost)? The answer is usually either a balanced price, meant to maximize revenue, or some sort of price discrimination (ie. business users pay more, home users pay less.)

2) Companies other than the telephone monopolies have dark fiber: these companies (or their creditors, who now are the proud owners of the fiber) should be, and probably are, perfectly willing to degrade the telcos' pricing power if they can make any money. The fact that they don't is because they *can't* make any money... for two reasons: (a) the cost of equipping the dark fiber is higher than the revenue they could generate from the few people who want broadband, and (b) the excess dark fiber at the telcos means they could push their pricing down to squash any new entrant (typical of industries that require high up-front investment: once the sunk cost is sunk, competitive threats are met by temporarily lowering the price to just above marginal cost, making it untenable for new.

Farming and steel are bad analogies: the US government pays subsidies to keep those industries alive in the US because they are strategic and the workers are influential voters. If there were no subsidies, those industries would migrate to where they belong: in countries where the cost of labor is lower.

Oil is a bad analogy for other reasons (as was the gold rush analogy in the article): this is a low variable-cost industry with a low fixed-cost as well, but restricted by the luck of having the natural resource in the first place. So, no member of OPEC produces *no* oil, the cartel simply tries to lower overall production to keep prices up. This would be analogous to *each* of the dark fiber owners agreeing to only use 5% of their fiber, to keep prices up. This is clearly not what is happening: some fiber owners are using none, while some are using substantially more than 5% (if 5% is the average, this has to be true.)

A good analogy is the railroad price wars of the 19th century. Early on, high up-front investments led to little competition, good pricing power and large margins and very profitable businesses. The early successes led to a large inflow of investment capital and the building of competitive railroads on the same routes (sometimes with tracks laid parallel to each other.) This led to price wars where both roads would charge just above marginal cost, not enough to cover fixed costs. While this made some economic sense, it led to the eventual demise and/or consolidation of one of the roads. These types of industries are generally considered natural monopolies.

Re:No way

[fhwang]

It's not because the phone companies are evil. It's because they're big, and fat, and it's way too easy for them to perpetuate the status quo.

Have you ever worked at a big company? I've worked at a few, and my personal experience is that in really large companies (say, more than 1000 employees) this very particular organizational rot sets in ... When the people making the decisions are so removed from their customers, they just stop caring. And if there is no competition to make them care, they'll just get fat and sleepy, and their customers will fall behind.

Residential DSL is the perfect example. Here in NYC, Verizon owns the phone lines, so all residential DSL has to go through them. In theory, they're supposed to allow equal access to all res-DSL companies, whether they're Verizon residential DSL or their own competition.

But I know dozens of people here who have DSL -- and nobody I knew was able to get DSL from a company other than Verizon. More than one person told me they tried to go with a smaller company, but the installation experience was really difficult: The other company couldn't do anything 'til Verizon flipped that switch, and somehow non-Verizon customers seem to get lower priority than Verizon customers. Curious, that.

A company doesn't have to be evil to screw you. Often, complacency is enough.

What's hindering broadband in the US?

[tamnir]

Japan has been laying dark fiber for over 10 years now. Many were laughing at them during that time, but today, we have a 100 Mbps fiber internet connection coming right into our kitchen! For something around 100$ a month. Ok, not super cheap yet, but affordable, specially if you share it.

Five years ago, the top for end users was still 56 kbps modems, that was just the begining of ISDN. Pretty impressive evolution.

Now question: if dark fiber is there, why is it that you still can't get decent DSL internet connection in the US? What's hindering the development of broadband there?

Dark Fiber gaffe or proper planning?

[Alomex]

If you were burying water mains and other city services into a new house subdivision nobody would be surprised about the city buring enough capacity for a 100 houses, even though only one has been built just yet.

Most of "dark fiber" articles out there fail to see the same rationale behind the large amount of dark fiber out there. This is proper planning. Network traffic has been doubling every two years or so, this means that 90-95% dark fiber would last you about 6-8 years.

This is perfectly sensible. In fact, if we had to rebury fiber within 6 years of paying billions to rip open downtown Manhattan I would fire my provisioning manager.

Re:No way

[Uhh_Duh]

Not half as stupid as your complete lack of information on this subject.

The key point you seem to not understand is that "Lack of bandwidth" no longer drives this market. There's more than enough bandwidth to go around with the leftover from the dot-com boom. Did it ever occur to you that maybe the reason they're not lighting up the fiber is that it's simply not needed right now?

Lighting up Fiber doesn't make bandwidth cheaper in this market since there's no demand. In reality, excess fiber would make bandwidth more expensive due to the increased overhead of having to maintain equipment and staff that aren't doing anything. Also remember, there's more to bringing bandwidth to the home or business than having fiber within a mile of the door.. The cost to trench it in and install the equipment, even if you're tapping from a short distance, is substantial -- well beyond the reach of any consumer or small business.

Furthermore, your arguments regarding anti-competitive behavior are even more ridiculous. If there's one industry where being a monopoly is a massive disadvantage, it's telecom. The Bells get screwed DAILY by the tariffs in place by the FCC (I don't have sympathy for them, they dug themselves into that mess) but business is NOT easy for them. The small-guy is at every advantage in this industry. If the big boys own the lines and the little guys want to use them, the FCC says they have to let them -- even if it means the big-boys taking a financial loss on the deal.

Sorry.. but you have much to learn about the telecom world before you open your mouth on the subject again.

Re:Proof of monopolies...

[scoove]

Cheaper technologies, like wireless, might well leap ahead of fiber in the race to more bandwidth.

This is what we deal with smack in the middle of fly-over-country USA.

While I've negotiated for over a year with several idle fiber network owners (who still expect rates greater than what the ILEC charges, yet requires me to sink hundreds of thousands of my own capital to build out their fiber to markets where it's usable), we've resorted in nearly every case to tower construction, licensed microwave DS3 deployment, etc. (Funny, we're cheaper *and* faster than fiber, yet have more than enough capacity to feed small towns).

I've argued the concept of "sunk cost" until I've been blue in the face - no good. Many of these guys came out of ILECs and have a fantasy about "price = cost + 40%", rather than understanding "price = what market will bear."

Blame it on too many laid-off Bellheads going to work for network companies.

I think most telcos think of the fiber network as their backbone, and they don't really market it as a service for business.

I think it's more of a "we won't sell anything less than our cost+high margin price." Funny though, nobody's taking.

Perhaps the little wireless companies that are growing strong will have some nice post-bankruptcy networks to buy penny on the dollar.

*scoove*

Re:Who owns the fiber

[Casca]

The fact that your post was modded +5 Interesting says something about the readers here at slashdot these days.

So how does the ownership of these lines pass on? Can just anybody take the existing lines, plug in, and make use of them - or do they have to be bought?

You're kidding right? Someone always owns everything, nothing of value is unowned. What, you think when someone goes out of business they just lock the doors and walk away, leaving a building full of inventory, office furniture, and whatever intact? Typically they go bankrupt, filing chapter 7 (liquidation). This means all of their assets are sold off and the creditors get the proceeds. This means someone is buying their assets, and dark fiber would be considered an asset by most. So no, you can't just use some defunct company's dark fiber.

If there were one large company that could buy out and connect most these unused lines, they could probably make something out of them. Since they're just sitting unused, I'd imagine it wouldn't cost too much to buy ownership

It was exactly this sort of thinking that put so many companies out of business to begin with.

Actually it shows they were trying to save $

[Ungrounded Lightning]

This article is just a reminder of how wasteful people were back in the DOT COM boom days. I'm sure that stories like these can be run in many major US cities.

Actually, this article is just another case of the media being sensationalist while either completely misunderstanding the situation or deliberately misconstruing it to hammer it into the current propaganda "template".

Virtually all of the fiber is SUPPOSED to be dark at this point. It's a side-effect of minimizing cost. Consider:

If you're going to do, say, a national fiber backbone network, you have to run a fiber loop around THE WHOLE COUNTRY, plus runs back-and-forth to hit all the major non-edge cities. As you go you lay conduit in the trench, pull fiber through the conduit, and splice it.

The cost of the fiber is NOTHING compared to the cost of the trench. You can put a WHOLE BUNCH of fibers in a single jacket, so the cost of the pull is the same. The cost of the splices is non-trivial and part of it is per-fiber-spliced, but it's STILL tiny compared to the trench.

So, how many fibers are you going to pull and splice?

It takes two fibers to make a section of the link, one for signals each way. That pair of fibers can carry (at the current top-of-the-line rate) about ten billion bits per second. That's 129,024 simultaneous uncompressed phone calls, or over a gigabyte/second of data traffic.

Now suppose you were only planning ahead for a couple years, and figured one pair would be enough. So you only buried one pair. And you got enough customers signed up IN ADVANCE to just about fill it. And you went to hook it up and found that somewhere between SF and LA there was a break. Are you going to dig up a third of the west coast again to fix the break? Of COURSE not! You're going to bury EXTRA FIBERS in the first place, and use a spare fiber. But suppose you have only one spare pair and your main fiber is full - that's 50% dark fiber! CNN Headline News screams "Half the fibers in the country are dark! Oh, the waste! Oh the horror!"

But do you, as the visionary building a network, think that the traffic is NOT going to increase in the future? If it doubles next year, do you want to light up another fiber? Or DIG ANOTHER TRENCH?

So of COURSE you spend a few extra percent up front. You bury a BIG BUNDLE of fibers. (You also bury a few extra conduits, so you can pull more, or rent-or-sell one to some OTHER networking upstart who wants to pull his own fiber, once you're safely established.)

So you're going to have a bunch of extra fibers. But how many do you light up? Answer: As few as possible. The boxes to light them up are NOT cheap. (Repeaters aren't muchFigure 1/8 million for a minimal TDM only box, over a meg for for a fully-loaded router.) But (unlike digging trenches) they are subject to Moore's Law improvements. Wait 18 months and your suppliers can get you twice the bits for the buck. So you buy expandable boxes (again to save costs later) but leave most of the slots empty.

And now you have most of the fibers dark, until the traffic expands enough for you to buy more cards and shove 'em into the boxes to light up the rest of the fibers.

So you have MOST of your fibers dark. And even reserving a few for spares you can light up most of 'em with paying traffic. But HOW MANY should you have?

The common wisdom at the time was that the Internet bandwidth needs were growing by a factor of 10 per year, and would continue that way for a while. If you have 95% dark fiber now, (and the bright fiber is at capacity), in one year you'd have half of it lit, and in another three months you'd hit the wall, and be frantically throttling links, upspeeding them with new technology, and getting out the cable-pullers and trenchers again. The bandwidth glut becomes a bandwidth crunch.

Turns out 10x/year was a myth, based on the explosive growth for the first couple years after the Internet was opened up to general users. The actual number is closer to 2x - which means today's 95% dark fiber means we don't have to get out the pullers and trenchers for a bit over a decade - and maybe longer if we go to higher speed over existing fibers.

Same situation in the metro networks - except that you're talking about digging up ALL THE STREETS OF ALL THE CITIES, TOWNS, AND VILLAGES. Then doing it again in a few years if you didn't lay enough up front.

It just makes you think.... How much stuff is out there that is just undocumented? How much wasted technology is out there that will never be found.

This is well known in the industry. It's just that the media are clueless.