You won’t have broadband competition without regulation

By Felix Salmon
February 21, 2014
Tyler Cowen isn’t worried about the cable companies’ broadband monopoly.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Tyler Cowen isn’t worried about the cable companies’ broadband monopoly. His argument, in a nutshell: if you can’t afford broadband, that’s not the end of the world: you can always go to the public library, or order DVDs by mail from Netflix. And if the cable companies’ broadband price is very high, then that just increases the amount of money that alternative broadband providers can potentially make in this “extremely dynamic market sector”. Indeed, he says, if regulators were to force cable companies to decrease their prices, then that would only serve to decrease the amount of money that a competitor could make, and thereby lengthen the amount of time it will take “to reach a more competitive equilibrium”.

The first big thing that Cowen misses here is television. Cowen knows that there’s more to broadband than watching movies on Netflix, but what he doesn’t really grok is that there’s more to Netflix than watching movies on Netflix. Netflix has moved away from the movies model (which was a constraint of the DVDs-by-mail model) to a TV model. And that makes sense, because Americans really love their TV. They love it so much that cable-TV penetration is still substantially higher than broadband penetration. As a result, any new broadband company will not be competing against the standalone cost of broadband from the cable operators: instead, they will be competing against the marginal extra cost of broadband from the cable company, for people who already have — and won’t give up — their cable TV.

If you’re a cable-TV subscriber, the cost of upgrading to a double-play package of cable TV and broadband is actually very low; what’s more, there’s a certain amount of convenience involved in just dealing with one company for both services. The result is the barriers to entry, in the broadband market, are incredibly high. Cowen talks about pCells and Google Fiber, but really they prove my point: pCells are untested technology which would surely cost a mind-boggling amount of money to roll out nationally, while it’s taking even the mighty Google a huge amount of time and money to bring its own broadband service to a relatively small number of mid-size cities.

What’s more, all of that effort is redundant and duplicative: we already have perfectly adequate pipes running into our homes, capable of delivering enough broadband for nearly everybody’s purposes. Creating a massive parallel national network of new pipes (or pCells, or whatever) is, frankly, a waste of money. The economics of wholesale bandwidth are little-understood, but they’re also incredibly effective, and have created a system whereby the amount of bandwidth in the US is more than enough to meet the needs of all its inhabitants. What’s more, as demand increases, the supply of bandwidth quite naturally increases to meet it. What we don’t need is anybody spending hundreds of billions of dollars to build out a brand-new nationwide broadband network.

What we do need, on the other hand, is the ability of different companies to provide broadband services to America’s households. And here’s where the real problem lies: the cable companies own the cable pipes, and the regulators refuse to force them to allow anybody else to provide services over those pipes. This is called local loop unbundling, it’s the main reason for low broadband prices in Europe, and of course it’s vehemently opposed by the cable companies.

Local loop unbundling, in the broadband space, would be vastly more effective than waiting for some hugely expensive new technology to be built, nationally, in parallel to the existing internet infrastructure. The problem with Cowen’s dream is precisely the monopoly rents that the cable companies are currently extracting. If and when any new competitor arrives, the local monopolist has more room to cut prices and drive the competitor out of business than the newcomer has.

In other words, the market in delivering broadband to the home is pretty much the opposite of the international text-messaging market which was disrupted so effectively (and so profitably) by WhatsApp. The initial impetus for WhatsApp came in Europe, where lots of people want to communicate with their friends across borders: from Germany to Austria, say, or from the Netherlands to Belgium. Text messaging across borders is expensive, both to send and to receive, and WhatsApp used those phones’ existing internet connectivity to be able to provide a better service at a price of zero. Since the mobile operators weren’t willing to bring their international text-messaging prices down to zero, they simply lost tens of billions of dollars’ worth of text messages to WhatsApp and other internet-based messaging services.

In broadband, by contrast, it’s the cable operators who could, if they wanted to, bring the marginal cost of broadband down to zero. (There’s no reason, in principle, why they can’t provide broadband for free to anybody with a cable-TV subscription.) Meanwhile, any would-be disruptor, needing to repay a massive capital investment, is going to have less ability to slash prices than the incumbents do.

So don’t count on competition to bring down prices in the broadband space. This is an area where the regulators — and only the regulators — can really be effective.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
15 comments so far

An FCC rule like this would take years to institute and – after court appeals – become final. Best case is at least three years, maybe five or more.

Anyone got any ideas on how to keep competition for content – Netflix TV, etc. – alive in the meantime?

Posted by dweightman | Report as abusive

You’re letting Cowen off way too easily. His understanding of technology is at the high school level, and his dismissal of the importance of broadband to the lower income strata is typical of economists (just use the library? really? you mean, after they work their two jobs?).

The very fact that he uses pcells as a reason to let the broadband buyers go about their business unregulated should be a tipoff. Not only is the Perlman scheme untested, it hasn’t even been vetted by any outside technologists. There has been no public non-superficial description of how it’s supposed to work, and Cowen thinks we should just accept it as our savior against cable companies (as if the U.S. wireless carriers, who are among the most expensive in the world, would be any less customer-focused).

And while it’s great that Google wants to provide its fiber service in 34 cities, keep in mind that could take five years, and, it’s only 34 cities. What do we do in the rest of the country?

Building out two wired networks is a waste of capital. To wire even just the urban areas of the U.S. with fiber would cost tens of billions of dollars. Even though Google is proving that investment would be profitable at $79/month, it doesn’t mean it would be profitable for two fiber service providers in each area, as there would be less customers on each network.

The best solution would be for municipalities to pay for installing a fiber network. They don’t have to operate it, but they should own it. They have access to the lowest cost of funding (tax-free debt), which would distribute the burden of funding the network over the user base, not a handful of giant companies. The local governments could put the operation of the networks out to bid for 5-7 year periods, and decide whether they want to pay the operator and provide access for free, or let the operator charge subscribers a pre-determined amount.

High speed (and I don’t mean 3 Mbps, which is what VZ and ATT call it in many areas) internet access is a basic necessity for 21st century life, just like electricity and street lights and schools. Just letting the market take care of providing it has proven that it won’t work – we have a second world broadband infrastructure, and pay the most in the world for it (kind of like our health care system). The incumbents have had at least 15 years to get it right and have demonstrated they have no interest in providing the best possible service at the lowest possible cost, but rather they offer as little value as they can get away with, while charging as much as possible. If they don’t want to be regulated, they need to face competition from their would-be customers, and the only way for those customers to accomplish that is via their local governments (maybe Cowen will suggest that residents in under-served communities go to their local coffee shops and start Kickstarter campaigns to build networks).

Those who oppose regulation need to be reminded that cable companies and the wireless and wireline providers have all been granted psuedo-monopolies for their service.. Also, in most cases, they have been given right-of-ways and easements, at no cost, to install their wiring. The wireless carriers are granted virtually exclusive use of spectrum (while maybe anyone can try to install fiber in their neighborhoods, nobody without a license can offer 4G service, and those licenses are very finite). In exchange for all of these public assets, those service providers should be forced to provide a public service, and not just extract rents.

Posted by KenG_CA | Report as abusive

>This is called local loop unbundling<

The wikipedia page linked to, mentions ILEC. ILECs, of course, implies CLECs.

Perhaps you could write up a piece about the rise and fall of CLECs in the late 1990′-early 2000′s. (Remember Covad? Speakeasy? Sonic? MegaPath?)

Posted by dtc | Report as abusive

Shorter Tyler Cowen: “Sure, natural monopolies are a concept taught in Econ 101 classes everywhere, and economists of all stripes have acknowledged for decades that utilities are the canonical example of this type of market structure, but my ideological hatred of government and regulation is so strong that I will now proceed to ignore what I know and kick up a bunch of dust to try to obscure and confuse the issue.”

Posted by Auros | Report as abusive

@KenG_CA you and I are usually on the opposite side of just about every issue but I’ll tip my hat to your municipality owned fiber back bone. I can’t imagine a public investment with a better return on investment… I’m lucky I live in a college town but every first world resident living inside city limits deserves +20MBPS for $20/month.

Posted by y2kurtus | Report as abusive

Tyler Cowen is right but so is Felix. In a natural monopoly you get innovation (Google “Ma Bell”) but also you get barriers to entry. However, you do get lower prices (my cable bill in the US is less than fiber optic and gets the job done for productive work as opposed to watching movies). But the potential gains that come from being a monopolist are sufficient for competitors to want to come into the sector–if regulators allow it–and this creates innovation. A good book on this (and the AT&T breakup) is the 1987 book by economist P. Temin.

Posted by RayLopez | Report as abusive

y2k, thanks, but I wouldn’t say we are usually on the opposite side of every issue, I think it’s closer to 50-50. If it was more like 90-10 you would hear from me more often. :-)

Posted by KenG_CA | Report as abusive

Deftly skewered, Felix. And Ken G., thanks for your equally insightful logic & comments upthread.

Posted by crocodilechuck | Report as abusive

Felix is not right about EU and cross border texting. Roaming is heavily regulated in EU and all cross country SMS cost 8 cents (will fall to 6 in July).
That means that you often pay much more for local texting than cross country texting.

Posted by simsimsim | Report as abusive

This column is severely factually-challenged. Granted, Tyler Cowan doesn’t understand broadband technology in any particularly deep and detailed way, but he gets the key aspect of that technology better than Felix Salmon does. Broadband is a dynamic technology, and its markets are built on a race between competitors to provide services that are faster and better year after year. American firms have invested $1.2B in wired and wireless broadband networks because this is a dynamic market.

Europe does not provide broadband cheaper than the US does when we correct for the two main cost factors, speed and distance, and include the effects of subsidies and taxes. In fact, only half of European residences have access to any kind of cable service, while close to 95% of American residences are passed by cable.

Local loop unbundling is a notion that applies to the telephone network, a system in which every residence has a unique strand of wire connecting it to a phone company office, but it doesn’t work for cable TV networks because 100 – 400 homes share a common wire. You can’t just disconnect a wire to the cable company to connect a given consumer to an alternate ISP. Unbundling cable would require access at cable company switch and close coordination between the cable company and the ISP. Even Europe doesn’t mandate that.

Traditional telco unbundling is one of the things that holds Europe back. American telcos are shortening their local loops by building neighborhood nodes with fiber backhaul. Europe can’t build these without coordinating a move for a half dozen minor ISPs and providing real estate for them in every node. This is why European regulators want to adopt the American system.

American networking companies are installing 19 million miles of fiber optic cable every year in order to speed up their networks. Verizon is testing LTE running at 80 Mbps and plans to use this to provide a competitive broadband alternative in rural and exurban areas. Europe can only dream of the kind of pervasive LTE we have here.

It’s always a good idea to do a little research before expressing opinions and attacking those who understand the issues better than you do. Here’s a start: http://www.itif.org/publications/whole-p icture-where-america-s-broadband-network s-really-stand & here’s something more recent: http://www.aei.org/outlook/economics/inn ovation/internet/the-european-unions-bro adband-challenge/

Posted by RichardBennett | Report as abusive

The economics of cable/internet services are pretty complicated and changing rapidly. Wouldn’t it be at least approximately correct to say that cable companies have been selling TV and that internet has been gravy on top of that? And maybe phone service as well? And the cable TV business seems to have been trapped into a model in which the content providers keep increasing the number of channels and ram them down the throats of the cable companies, the end-user ultimately paying. But now content is shifting to the internet, so things may flip to prices based more on internet. If cable TV dies, this will not make internet provision cheaper.

Posted by skeptonomist2 | Report as abusive

Let’s try this again since my comment hasn’t posted yet. This time without links.

This column is severely factually-challenged. Granted, Tyler Cowan doesn’t understand broadband technology in any particularly deep and detailed way, but he gets the key aspect of that technology better than Felix Salmon does. Broadband is a dynamic technology, and its markets are built on a race between competitors to provide services that are faster and better year after year. American firms have invested $1.2B in wired and wireless broadband networks because this is a dynamic market.

Europe does not provide broadband cheaper than the US does when we correct for the two main cost factors, speed and distance, and include the effects of subsidies and taxes. In fact, only half of European residences have access to any kind of cable service, while close to 95% of American residences are passed by cable.

Local loop unbundling is a notion that applies to the telephone network, a system in which every residence has a unique strand of wire connecting it to a phone company office, but it doesn’t work for cable TV networks because 100 – 400 homes share a common wire. You can’t just disconnect a wire to the cable company to connect a given consumer to an alternate ISP. Unbundling cable would require access at cable company switch and close coordination between the cable company and the ISP. Even Europe doesn’t mandate that.

Traditional telco unbundling is one of the things that holds Europe back. American telcos are shortening their local loops by building neighborhood nodes with fiber backhaul. Europe can’t build these without coordinating a move for a half dozen minor ISPs and providing real estate for them in every node. This is why European regulators want to adopt the American system.

American networking companies are installing 19 million miles of fiber optic cable every year in order to speed up their networks. Verizon is testing LTE running at 80 Mbps and plans to use this to provide a competitive broadband alternative in rural and exurban areas. Europe can only dream of the kind of pervasive LTE we have here.

It’s always a good idea to do a little research before expressing opinions and attacking those who understand the issues better than you do. Here’s a start: Google “The Whole Picture Where America’s Broadband Networks Really Stand” & “The European Union’s Broadband Challenge”. Europe is not broadband utopia and unbundling doesn’t help.

Posted by RichardBennett | Report as abusive

@RichardBennett, “American firms have invested $1.2B in wired and wireless broadband networks because this is a dynamic market.”

The telcos love to brag about the large sounding amounts they invest in their networks, but $1.2B is not a lot of money when you consider there are over 300 million people in this country. Compare that number with their revenue, which is in the hundreds of billions of dollars, and it seems pathetically small.

“Broadband is a dynamic technology, and its markets are built on a race between competitors to provide services that are faster and better year after year”.

Verizon has not increased the speed of their DSL service, nor offered FiOS, at two of the homes where I manage the communications service. In over 10 years. It would not cost a lot relative to their monthly fee to increase the speed of those DSL connections, but they choose not to make that investment. And they have halted expansion of their FiOS service. There is no race to provide services that are faster or better, or if there is a race, it is in increasing advertising budgets for marketing the fantasy that they offer “high speed” broadband.

LTE cannot be a serious alternative to wired broadband in urban and suburban areas, as it is a shared medium, unlike DSL or fiber. Also, the main attraction for LTE for the carriers is that their data caps are low and they will be easily exceeded as video distribution shifts more to the Internet.

Posted by KenG_CA | Report as abusive

KenG – that $1.2 billion number is clearly not correct. Any single major company in wired or wireless broadband markets – AT&T, Verizon, Comcast, Time Warner Cable, etc. – spends more than that on network capital expenditures every year. To take just 1 example, AT&T’s total cap ex spending is about $20 billion per year – http://www.lightwaveonline.com/articles/ 2013/04/at-t-trims-capex-estimates-for-2 014-and-2015.html

Posted by realist50 | Report as abusive

Realist, I knew that number was ridiculously low, but I couldn’t let it go.

Still, $20 billion in capex for a telco with revenue of $128 billion is not special. And I would bet a disproportionate amount is spent on their LTE network. I doubt much is spent expanding their fiber footprint or upgrading their DSL service.

Posted by KenG_CA | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/