Monopolizing bandwidth

By Felix Salmon
February 17, 2014

Paul Krugman makes a simple but powerful point about Comcast’s acquisition of Time Warner Cable:

One puzzle about recent U.S. experience has been the disconnect between profits and investment. Profits are at a record high as a share of G.D.P., yet corporations aren’t reinvesting their returns in their businesses. Instead, they’re buying back shares, or accumulating huge piles of cash. This is exactly what you’d expect to see if a lot of those record profits represent monopoly rents.

Broadband is the area which Krugman, and most other opponents of the Comcast-Time Warner tie-up, are most worried about. It can’t be a good idea to give a single company 37% of the market in broadband, especially when its real monopoly power would be much stronger still:

The reason this deal is scary is that for the vast majority of businesses in 19 of the 20 largest metropolitan areas in the country, their only choice for a high-capacity wired connection will be Comcast. Comcast, in turn, has its own built-in conflicts of interest: It will be serving the interests of its shareholders by keeping investments in its network as low as possible — in particular, making no move to provide the world-class fiber-optic connections that are now standard and cheap in other countries — and extracting as much rent as it can, in all kinds of ways. Comcast, for purposes of today’s public , is calling itself a “cable company.” It no longer is. Comcast sells infrastructure subject to neither competition nor a cop on the beat.

The argument from Comcast is, essentially, that it doesn’t matter whether it has a national monopoly, because it (and Time Warner Cable) already have local monopolies. If individuals and businesses don’t have any choice of broadband providers right now, then what difference does it make if the existing providers consolidate?

The argument does have a little bit of merit, if you believe that the main reason not to have monopolies is to encourage competition. But take a step back, and it’s abundantly clear that the US has something approaching a national broadband crisis on its hands.


In comparison with the rest of the developed world, the US has slower broadband speeds and higher broadband prices than just about anybody. When you do find exceptions, they always turn out to be cases of a very clear monopoly: Carlos Slim more or less owns broadband in Mexico, for instance, while a company called Southern Cross controls all of the bandwidth into New Zealand.

What’s more, in cases like Mexico and New Zealand, the rule of supply and demand at least still obtains. Broadband prices are high — but in large part that’s because the supply is constrained. The supply is constrained mainly because the monopolist sees no particular reason to increase it: they’re already charging monopoly prices, which means that they wouldn’t make more money by providing better service.

The US, by contrast, is unique in that it has very high broadband prices and an abundance of bandwidth. The country as a whole — or at least its urban centers — has no shortage of bandwidth at all. But if you want to connect your home or business to the major internet backbones, the cable-company gatekeepers will charge you an arm and a leg for doing so.

Farhad Manjoo has the explanation for why this should be. Internet service is very cheap for the cable companies to provide, and it’s also price-sensitive: if you reduce the price, more people will sign up. As a result, the cable companies would make more money from their broadband offerings if they reduced the price. So why don’t they? Because right now, 91% of Americans with broadband also have cable TV (I think, I can’t find the link for that right now), and the cable companies make their real money from TV, not broadband. The cable companies therefore have every incentive to price broadband as high as possible, so as to make the marginal extra cost of getting TV as well as small as possible.

In the US, cable TV rates are very high; as such, the best way to prevent cord-cutting is to ensure that broadband rates are also very high. That’s bad for broadband adoption, but it’s reasonably effective at keeping people paying very large sums for TV every month. In other words, high broadband rates are a bit like most newspaper paywalls: they’re not so much a way of making lots of money themselves, as they are a way of persuading you to pay lots of money for something else. (Physical newspaper delivery, or cable TV.)

If Comcast is allowed to buy Time Warner Cable, that model won’t change — but it will be reinforced. The cable companies will continue to price broadband at uneconomically high rates, in order to protect their cable TV cash cows. And as Krugman notes, they will have essentially no incentive to improve their own broadband infrastructure, since providing high-quality broadband is not how they make money. Instead, they will just continue to extract monopoly rents, which is good for their shareholders, but bad for everybody else.

There isn’t a market solution, here: there’s only a regulatory solution. The US government regulates the amount that the post office can charge, so that everybody has access to the mail; it also regulates the maximum amount that phone companies can charge for basic landline telephone service. Both of those regulations are beginning to look increasingly anachronistic, in an era where the internet has replaced both mail and telephony. But the obvious regulatory response — to mandate that utilities provide universal access to low-price, high-quality broadband — seems as far away as ever. If Comcast is allowed to buy Time Warner Cable, the current model will become even more entrenched. And the USA will slide ever further backwards in the global connectivity race.


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

Let’s think about this for a second – internet is pure profit. Comcast takes 100% of internet revenues, closer to 50% of cable revenues. So no, cable is not more profitable. Not even close.

Posted by billyjoerob | Report as abusive

Have the government use ED to buy the last mile of all cable/fiber. Offer anyone access to that last mile at a reasonable rate that will also allow for the last mile to be upgraded as technology advances.

Posted by Zdneal | Report as abusive

And while individuals are local and have little choice, businesses are often not and have some choice over where to provision which a larger monopoly would reduce.

Posted by MyLord | Report as abusive

Zdneal, that is a great idea, but unfortunately, 20 states have laws limiting or prohibiting local and regional governments from offering or supporting internet service (  /02/isp-lobby-has-already-won-limits-on -public-broadband-in-20-states/). In fact, a law was introduced just a few weeks ago in the land of Oz that would prohibit municipalities from investing in or providing broadband service for their residents. This bill was not introduced by residents of Kansas, but by the cable industry, of which Comcast is a leading member. It is apparently in response to Kansas City working with Google to bring their residents internet service that rivals what people in South Korea can get.

Posted by KenG_CA | Report as abusive

That chart would be a bit more compelling with Australia also on it since it too has a low population density. The US and Canada are most decidedly not like those other countries when it comes to spreading wired infrastructure and it is not surprising to see things different.

That said I agree with the overall point and the emphasis on the pernicious nature of infrastructure monopolies entirely, but you want to make sure you have all your argumentative ducks in a row when taking on a lobbying juggernaut like Comcast.

Posted by QCIC | Report as abusive

@QCIC, I fear your point regarding population density is specious, because the US also compares unfavorably in like-for-like comparisons of metro areas with similar population densities (e.g. greater Seoul vs. Los Angeles, or Hong Kong vs. Manhattan).

Further, rural areas in northern Sweden or Finland, where the population density is lower than all but a handful of counties in the lower 48 states, also enjoy far faster broadband at much cheaper rates.

Posted by David4321 | Report as abusive

Internet is 100% profit? Where are you getting that idea? The marginal cost to add _a_ customer to your internet service may be very low, but the cost to add a thousand, or a million — or to increase the bandwidth provided to existing thousands or millions — is quite steep, particularly if you want to serve rural areas.

As David4321 says, the population density thing is a canard, if you’re talking about serving the metropolitan areas. If we’re going to mandate ultra-high-speed service for *everyone*, including people living out in the rural hinterlands, there will be issues with figuring out who’s going to pay for that, but they’re entirely surmountable.

Posted by Auros | Report as abusive

How come there is zero mention of telcos as competing providers of broadband?

Posted by realist50 | Report as abusive

Realist, because unless you can get FiOS, most telcos don’t offer what is considered 2010 broadband. The fastest speed DSL that VZ offers me is about 3% (3 Mbps) of the max speed of my local cable company. The highest speed VZ offers to the street where my office is located is 7 Mbps (technically it’s 10, but 7 is as fast as you can get if you want static IP addresses). And it’s not that the technology doesn’t allow faster DSL, but rather than VZ and ATT are hoping their subscribers will be happy with severely capped LTE, so they won’t upgrade their late 90′s era DSL to a faster DSL. They avoid competition even more than the cable companies.

Posted by KenG_CA | Report as abusive

Well, the practical differential between DSL and cable is smaller than the headline numbers would lead you to believe. The speed I get drops during peak usage hours, when a lot of other people on my local loop are using Netflix streaming and the like.

But it’s true that unless you’re practically nextdoor to a CO, the speed on DSL is going to be slower. OTOH, I used to have a situation where I could reliably get 5-6 MB on DSL b/c I happened to be within a thousand feet of the nearest CO. The cable I have now definitely tops out at speeds much higher than that, but I don’t think it reliably does better all the time.

Posted by Auros | Report as abusive

This is the best article I’ve read on the subject. Thanks Felix.

Posted by SunnyDaySam | Report as abusive

“There isn’t a market solution, here” — Actually there is… the market cap of Comcast is about $150 billion. US federal government could buy it and spin out entertainment asset to one entity and infrastructure to another and lease infrastructure to service providers. The net cost would be less than 1 month of Fed bond purchases. Benefit would be a new business model in operation.

Posted by dbrodess | Report as abusive

Auros, almost all cable companies have been improving the speed of their networks, while DSL-based service providers have stuck with their 15 year old technology. In all of the homes that my extended family does not have FiOS (those houses are in NJ), cable performance, even during peak hours, is at least 5x that of the highest speed DSL offered.

As for distance, yes, DSL speed does drop with distance between the customer and the central office equipment, but that is why they can install remote terminals, similar to how cable companies install mini-head ends in each neighborhood. And if you are only 1000 feet from the CO, you should be able to get > 30 Mbps over DSL, if only your telco would upgrade to technology introduced in the last seven years or so.

The two large DSL providers have chosen not to upgrade their DSL networks, whether it’s because they hope customers will switch to LTE or they just don’t care. The cost of upgrading DSL to the latest technology can probably be paid off with less than three months of subscriber fees.

Posted by KenG_CA | Report as abusive

KenG, the cable companies haven’t increased the throughput speed to end-users since broadband first became available. They have only made improvements that allow more end-users.

Posted by Lycanth | Report as abusive

Paul Krugman’s argument about lack of incentives may be true for most companies, but simply is not the case for Comcast. From the beginning, Comcast has been run like a startup, reinvesting profits in its infrastructure, programming and new services. This argument is made more forcefully by former Ovation COO Chad Gutstein elsewhere. cast-time-warner-cable-bigger-doesnt-mea n-badder-1201103216/

The argument that most of the rest of the world has better broadband because they don’t suffer from local monopolies like Comcast is simply not true. Over and over, the rest of the world has government-backed monopolies that had slow speeds until the beginning of the 21st century. What changed? During the 90s, most countries had no cable, had no wireless, etc. but saw the need to upgrade their facilities — and in one fell swoop, surpassed the United States.

The place where Comcast truly falls short is in its support of Streampix. If the Net Neutrality lobby were to get its act together, this is clearly a situation where Comcast is prioritizing its video bits over others’ video bits. For years, Net Neutrality have had opportunities to establish this common-sense standard, but have instead chosen to support a loosely defined list of principles where advocates and critics alike can pick and choose the points they want to support their arguments. This approach is great for grabbing headlines but lousy for consumers.

Posted by connectme | Report as abusive

Our DSL tests at around 1 MB/s. Certainly not state of the art, and awkward for video (if there is anything at all going on at the same time). An extra $20/month for FIOS? Upgraded DSL would be fine, but of course Verizon is invested in their premium service.

Posted by TFF | Report as abusive

just another avenue for greed, and control, to manifest itself in Amerika….the land of the Korporate. while a docile electorate downloads another app and wonders why it doesn’t work so well.

Posted by rikfre | Report as abusive

The problem with this argument is that the facts contradict the claims of vast foreign superiority in broadband.

For a serious analysis of the real-world data on comparative broadband performance and cost, see: e-isps-gouging-us-broadband-users/

Posted by jsgreenfield | Report as abusive

To be contrarian here on the sentence “they will have essentially no incentive to improve their own broadband infrastructure” cable TV companies have invested something like $200 billion to upgrade from their old, 1-way analog networks to digital hybrid fiber-coaxial networks capable of delivering broadband (note they did this without any government subsidies). What was their incentive other than to innovate, satisfy customer demand and build their businesses? To the extent that the video industry only becomes more competitive going forward (Netflix now has over 33 million subscribers in the U.S. alone) wouldn’t they be expected to continue innovating in order to remain competitive?

I’m not really sure on what basis Krugman makes his statement?

Posted by BuzzyR | Report as abusive

To be contrarian here on the sentence “they will have essentially no incentive to improve their own broadband infrastructure” cable TV companies have invested something like $200 billion to upgrade from their old, 1-way analog networks to digital hybrid fiber-coaxial networks capable of delivering broadband (note they did this without any government subsidies). What was their incentive other than to innovate, satisfy customer demand and build their businesses? To the extent that the video industry only becomes more competitive going forward (Netflix now has over 33 million subscribers in the U.S. alone) wouldn’t they be expected to continue innovating in order to remain competitive?

I’m not really sure on what basis Krugman makes his statement?

Posted by BuzzyR | Report as abusive

There is so much wrong here that it’s hard to know where to begin.

On the one hand, we are approaching a national broadband crisis. On the other hand, we have an abundance of supply of broadband capacity. How can both be true?

Monopolies have constrained capacity because they have no incentive to invest. The US has plenty of capacity, but that’s still a product of monopoly? Again, aren’t these contradictory?

There are certainly some serious logic problems here, but then they’re also built on factually incorrect suppositions.

The chart is attributed to OECD (the same graphic is cited in articles all over the web, claiming the same OECD attribution), but is not locatable anywhere on OECD site, has no links to an underlying OECD report and methodology, and no indication of what timeframe it is supposed to reflect.

What we can say from the nice round speed numbers on the charts, as well as OECD’s stated methodology for other broadband reporting, is that OECD reports based on marketing offers, from a small sample of providers. They report speeds based on advertised speeds, without regard to what is actually delivered. They may also be reporting pricing before it is loaded with mandatory fees and taxes.

So presumably the reported data above reflects the best speeds and pricing found advertised, regardless of how widely or narrowly it is available in the given country, regardless of whether the speeds are actually delivered, and potentially without regard to whether the prices exclude “hidden fees.”

But, do any of the many people who have posted rants based on the above chart actually know how the chart was constructed, and by whom? Is anybody able to provide an actual citation to the source and methodology?

Then also, an underlying assumption: “Internet service is very cheap to provide.” This claim is based on what? Certainly not fact, and certainly not insight based on knowledge of the broadband business, because it’s dead wrong.

Internet service is very expensive to provide, not in terms of operating expense which critics myopically focus on because it suits their argument, but in terms of capital expense for unending capacity enhancement, and nowhere more than in the US, where per-user utilizations are much higher than the rest of the world, nearly 3x that of Europe, and have been growing at 40+% CAGR for many years, with no end in sight.

Anybody who is interested in understanding how the US really compares to other countries, including transparently sourced real-world data, and fully explained analysis, can find the actual facts here: e-isps-gouging-us-broadband-users/

Posted by jsgreenfield | Report as abusive

Even this dyed in the wool capitalist considers the TWX/CMCSA merger to be totally unacceptable on anti-trust grounds. My god the real cell phone companies are required to offer their own networks to competitors on competitive terms. I think there are 7 possible cell providers (sharing 4 real networks) where I live… I have 1 cable option and DSL is hardly broadband at all.

No freaking way this merger gets approved. And rather than just piss and moan about it on blog comments file a public comment with the DOJ when the case opens. Here is a link to their guidelines for horizontal mergers… this proposed transaction seems to violate at lest half of them… ines/hmg-2010.html

Posted by y2kurtus | Report as abusive

Also, New Zealand is thousands of kilometres away from, well, everywhere, so the physical reality of adding more bandwidth (i.e. laying additional submarine fibre) is non-trivial, and expensive.

Posted by mikee_eekim | Report as abusive


Ah, but they wouldn’t be providing broadband. They would be providing wires/fibers/cables (OSI Layer 1). The stuff that goes on the wires (OSI Layer 2+) would be provided by some sufficiently corrupt capitalist entity.

Posted by Zdneal | Report as abusive

I’m very happy to read this. This is the kind of manual that needs to be given and not the random misinformation that is at the other blogs. Appreciate your sharing this greatest doc.

Rattling superb info can be found on weblog . “I know of no great men except those who have rendered great service to the human race.” by Francois Marie Arouet Voltaire.

Looking forward to reading more!