Is there really a cable/mobile duopoly in America?

by on August 7, 2013 · 30 comments

This is the third of a series of three blog posts about broadband in America in response to Susan Crawford’s book Captive Audience and her recent blog post responding to positive assessments of America’s broadband marketplace in the New York Times. Read the first and second blog.

If Crawford’s mind, this is a battle between the oppressor and the oppressed:  Big cable and big mobile vs. consumers. Consumers can’t switch from cable because there are no adequate substitutes. Worst of all, she claims, the poor are hardest hit because they have “only” the choice of mobile.

Before we go deeper into these arguments, we should take a look back.  It was not long ago that we didn’t have broadband or mobile phones.  In less than two decades, our society and economy have been transformed by the internet, and we have evolved so quickly that we can now discuss which kind of network we should have, how fast it is, which kind of device to use, and even how the traffic should be managed on that network. The fact that we have this discussion shows the enormous progress we’ve made in a short time. Plus we can discuss it on a blogging platform, yet another innovation enabled the internet.

Defining Competition:  Economists vs. Lawyers

Economics and lawyers differ on how they define competition.  Economists define a competitive market which has many firms, homogeneous products, free entry and exit from the market, independence of decisions among firms, and complete information.  They define an oligopoly by the amount it differs from these factors.  Lawyers, on the other hand, have in mind a standard of evaluation and to what extent firms serves the public interest.  The legal definition is necessarily subjective because it takes into account a lawyer’s value judgments.

Even a cursory look at the American broadband market shows that it is complex.  The website of the National Broadband Plan has a wealth of data about broadband in the USA, including breakdowns of providers by zip code.  As of December 2012 there are 2,083 broadband providers, of which 1,618 offer basic broadband speeds of 3 Mbps; 1,018 offer broadband speeds of 6 Mbps, and 200 offer 100-megabit connections.[1] Check out the report of the number of providers by speed tiers. Thus the idea of a duopoly is hard to prove by the numbers. Further it is hard to call broadband a homogeneous product as it is delivered on at least 5 different network technologies, appears in many tiers, and is packaged in a variety of ways based on the user.

Whether it is difficult to enter or exit the broadband market may depend on the municipality, and to be sure, broadband providers are regulated. As for information about broadband providers, it gets better all the time. Not only is the National Broadband Map a useful tool, there are a many websites about pricing and consumer reviews. Needless to say, consumers have many outlets for information on broadband from the providers themselves.

Crawford is one legal scholar who believes the duopoly thesis, but others don’t.  University of Pennsylvania Law School professor Christopher Yoo observed, “There has never been more competition in the cable industry than today” on Jerry Brito’s podcast. His book The Dynamic Internet:  How Technology, Users and Businesses Are Transforming the Network makes a compelling case about the complex internet ecosystem and how a myriad of actors create the market.  No one company or industry emerges dominant.

Intermodal Competition

Crawford doesn’t believe there is competition between different types of networks, but the Organization of Economic Cooperation and Development (OECD) ranks the US as #3 in world for intermodal competition.  Even though the US is is a member of the OECD, it would be a stretch to call it a US-centric body.  This group coordinates the G-20, is based in Paris, and has as its primary activities data collection and analysis.  The notion of intermodal competition means that a consumer has a variety of networks to choose from:  DSL, cable, mobile, satellite, and Wi-Fi.

The idea of cable/mobile duopoly would mean that there are only two networks, each with just two firms; Comcast and TimeWarner for cable, and AT&T and Verizon for mobile.   It’s difficult for me to swallow this notion because when I visit or live in the US these are not my providers, and further, I know many people in different parts of the US who have other providers.

Personally I use a 4G mobile dongle as my broadband connection, and it accommodates fast downloading and uploading of video. In spite of ample wire line infrastructure in Denmark, already 7% of the population uses mobile as its primary source of broadband. In fact, I have never in my life subscribed to cable; there are just too many books to read. Russ Roberts of the Mercatus Center noted that he does not subscribe to cable because he and his three sons would spend the entire day watching sports.[2]  Many professionals I know don’t have the time to watch long format video, and for them, DSL does the job.  However I know plenty of people who love cable.   Furthermore I know literate, gainfully employed people who don’t care to spend their life on the internet, or if so, only sparingly.  If anything, my sense is that many people are overwhelmed by the choices of broadband networks.

Another wrinkle in the duopoly thesis is that satellite broadband  is available to 99% of Americans. This is important technology for much of the country is mountainous and not well populated.  Crawford scoffs at satellite broadband because it is “generally considered unsuitable for 21st century uses”, but it’s perhaps  because the only information she cites is For a more thorough discussion, see customer 24 reviews of ViaSat’s Exede  on DSLReports and reviews of 5 other satellite broadband companies.

Satellite broadband packages of 5-15 Mbps download start at $40/month[3]. There is a fee for equipment (for example a $10/month), or the equipment can be bought outright.  Satellite broadband is more than adequate for web browsing and email, the essential applications job hunting or health information.  Granted, it can’t be used to play video games and is probably not the best choice for VoIP applications such as Skype, but people use satellite broadband to watch Netflix (though satellite does reach data cap limits faster than wire line options).  See the demonstration of how fast a 20/6 Mbps satellite broadband connection loads in comparison to a fiber network.

There are power users of broadband for whom satellite is not the right choice.  They play massive multiplayer online games.  They run YouTube 24/7. They are active in peer to peer file sharing.  They have a set of needs, but their needs are not the same as grandmothers who use an iPad to play bridge, send email and check pictures on Facebook.  The market should be able to respond to different needs—and price points–without imposing one standard on everyone.

Many like to go about evaluating competition but counting the number of players in the industry. But if you live in the age of the automobile, it matters little to you that are are 100 horse & buggy companies.  Indeed a competition specialists might assert that if antitrust rules are written well enough, there is little need for industrial regulation. This is essentially the criticism of the European market where DSL makes up 75% of broadband connections.  This is the outcome of the 28 national telecom regulators counting the number of entrants that get to use the incumbent’s copper wires.  If you can get free ride on infrastructure, there is no need invest in something different.  Cable makes up just 15% of broadband connections in the EU, according to EU Cable, a trade association. The US has a more balance between DSL and cable technologies, enabling them to compete with one another.[4]

Overall, Europeans regret losing first place in mobile as the USA has taken a quantum leap in LTE. Thus many Europeans are pushing for a digital single market so that their homegrown web companies might better compete against Google, Facebook, Amazon, LinkedIn and the other American broadband-based companies that dominate the European landscape.  It’s for this reason the EU Vice-President wants to allow European cable and telco companies to merge, so they don’t have the inefficiencies of operating with the individual rules of 27 different countries.

Wired vs. Wireless

Crawford insists that every American should have two broadband subscriptions, one wired and one wireless.  She fails to realize that many of us don’t want or need both of these.  The Progessive Policy Institute published a report by Clinton Administration economist Everret Ehrlich called Shaping the Digital Age: A Progressive Broadband Agenda which notes  “Thus, while activists claim that only a high-speed, wireline connection will suffice, consumers are moving in an entirely different direction, toward wireless.  They are driven by their own needs and preferences, whether it is because they rent or move, because they prefer mobility and convenience, because they can accomplish whatever tasks they want to do on a mobile system, or for other reasons. Demanding that they have access to a wireline system in the name of ‘competitiveness’ is a waste of resources and an elitist substitution of planners’ preferences for a competitive market.”

Indeed many can do what they need to do on the web with mobile alone, and for them a wireline subscription is needless money spent on amusement.  Over sixty percent of America’s wireline broadband usage goes to entertainment.[5]   Yes, American music and movies are wonderful, but there are only so many hours in a day.  As my parents said to me about declining to upgrade their cable subscription, “We have enough movies.  We would rather play with the grandchildren.” People have different needs which are matched by different networks at different prices.  It’s not my place to tell people what kind of broadband they should have, nor is it Crawford’s.

Finance in the Cable/Telco Industry 

Whether we like it or not, financial markets are a part of the cable/telecom industry. They provide capital for infrastructure, and investors rely on capital gains to fuel their retirement funds. Crawford portrays cable and telecom companies as greedy, but  Valueline’s 2013 reports puts telecom services in the lower bottom half of all global industries for return on capital, 13.66%, placing it just above the publishing industry.

Crawford decries that the cable industry invested some 30% of revenues in 2001, but just 12-14% in recent years.  However this can be explained simply by the fact that 2001 marked a major investment in the DOCSIS innovation. That year CAPEX was triple the level of 1998. Shifting from analog to digital TV was a game changer for cable. Naturally after a big shift, these numbers will decline in following years.  However cable CAPEX jumped up again in 2006 and has stayed relatively stable since.  In general, the cable/telco industry invests 13% of sales in innovation, a higher percentage than other equipment industries.

Crawford’s claim about higher cable ARPU (average revenue per user) can be explained by the fact that cable providers now offer broadband and telephony in addition to pay TV. To focus only on revenue and not profit margin or capital expenditure does not tell the whole story.  From 1999 to 2009, Comcast’s return on capital tripled, but that amount was just 7%. Yes, a big company will have more revenue and larger dividends, but it will also have more costs.

Crawford further charges that between 2002 and 2012, AT&T’s dividend increased by 64%, while Verizon’s grew by 47%.  Again the answer is simple if you look at the history.  Between 2000-2002 the American carriers divested their assets in Europe and other regions. At that time, there were a different mix of companies (Bell South, Ameritech etc) which later merged into AT&T. Verizon had a similar M&A evolution.  When a sale of assets occurs in a publicly traded company, any profits will be returned to shareholders. Crawford implies that this money was derived from overcharging American customers, but it was not.  This just reflects corporate finance activities.

Crawford further singles out AT&T and Verizon for neglecting their wires and focusing on the more profitable wireless business. I doubt this because there are too many stakeholders holding AT&T’s feet to the fire, from the FCC to investors to savvy consumers on social media. AT&T closed 2012 with a net profit margin of 5.7%, hardly the stuff of a swindler.  Naturally this number will fluctuate; in 2011 it was 3.11% and in 2010 it was 15.98 percent. The point is that a telco will have many business lines, but it has only one stock that trades on the exchange.   Thus it has an incentive to manage all its business lines well to maximize its profit and share price.

Though there are various metrics to consider, it’s hard to make a case that telecom companies are fleecing their customers when one looks at the profit margins and the fact that investors have many other choices for industries which have higher returns such as software, aerospace, chemicals, pharmaceuticals and so on.  Perhaps most telling is the fact that internet companies whose businesses are built on top of broadband infrastructure (Google, Facebook, Netflix etc) are generally more profitable than the network providers themselves.  Not only do carriers effectively subsidize the leading internet companies with broadband infrastructure and data delivery, they also subsidize equipment such as handsets, modems, set top boxes, satellite dishes and so on.

The Internet Ecosystem

Broadband is not just about networks.  The complex internet value chain includes equipment providers, software providers, device manufacturers, content and application providers, and users.  Carriers are not the only actors, and their decisions are impacted by the participants around them.  We cannot overstate the role of content/application providers and device manufacturers.  In essence, these are the reason people get on the internet in the first place.  Therefore these groups have the ability to drive major economic change and innovation with their offerings.

While it can make for a good yarn that there is a some cable/mobile duopoly, the reality is that the internet value chain is too complex for any one or two players to exert extraordinary control.  To focus on one or two actors in the value chain is a static and monolithic analysis.  It does not allow for inevitable change and evolution which happens quickly with the internet.  I suppose Crawford would have sacrificed some of her book’s sensational appeal by allowing for greys instead of black and white.  The novelist has this license, not the social scientist.

The other economic force in the broadband marketplace is over the top competition, the services on the internet itself that compete with the carriers, such as Skype, WhatsApp and Netflix.  Skype managed to disrupt the global market for long distance.  WhatsApp caused SMS revenue to plummet 40% for some carriers.  Even Google’s Chromecast, a dongle which enables streaming YouTube and Netflix from Google to a digital TV, offers consumers a cable-free existence for only for $35. These competitive forces change the economics for networks and render the duopoly thesis even less valid. It’s not the number of players that creates competition, but the technological development.

These blog posts have reviewed America’s broadband in relation to the rest of the world, competing broadband technologies for the future, and whether certain firms exert extraordinary influence on the broadband marketplace.  With data from the OECD, FCC, Akamai  and my university, interviews with Americans, and my personal experience living in a variety of countries, I conclude that the American broadband market is competitive and robust.  Living abroad one comes to appreciate all the good things about America; broadband and the vast economy it enables are two of them.  If this is what a legal scholar considers duopoly, then I would like some more.

The final part of this series investigates a digital literacy program and how it can help those without an internet connection get online. It also addresses to what extent cost is a barrier to internet access.




  • Michael Elling

    It would help if the author pointed out early on in her discourse that mobile and broadband and the internet all developed out of 4 variations of “equal” or open access or net neutrality:
    1) dial 1 that drove voice competition and digitized our voice and enterprise data networks and drove “free calling.”
    2) cellular A/B interconnect/roaming that was extended to PCS carriers in the early 1990s from which 10 cent pricing soon followed and exploded usage and penetration (unlike in Europe which went the thumb-bound SMS route)
    3) Computer 2 which protected nascent ISPs in the early 1990s as they lawfully exploited a pricing loophole (aka flat rate dial-up across expanded LATAs) that the Baby Bells implemented to withstand the competitive WAN threat in the mid to late 1980s. Again, Europe didn’t have this benefit which is in large part why the internet scaled here.
    4) Wifi, or shared access, or infinite/nano cellular to not only drive broadband adoption but smartphone adoption as well.

    Extrapolate the resulting impact of those 4 policies and “voila” you get the miracle of the 1980s-90s that extended into the 2000s even as we remonopolized the mid and last mile. The price differential between retail and actual cost is now so great that those vertically integrated last mile providers will have a hard time sustaining their business models. We see that clearly with the Telcos. So even though the monopolists have managed to abolish or control 3 of the 4 equal access policies, it won’t help.

    If we don’t understand or see history for what it really was at that time, then we have no way of predicting the future. Network value develops from getting more people on them. The value develops sometimes at the core and sometimes at the edge. Horizontal layers and vertical boundaries need to become fluid. Marginal consumption and demand can only be served by competitive, agile service providers that scale horizontally to rapidly depreciate opex or capex in the lower, middle and upper layers. It’s time for a new business model. This debate is essentially between liberal statists, free-market libertarians and capitalist conservatives who at the end of the day believe in the natural monopoly fiction born conveniently 100 years ago to control information. Time to change that way of thinking as none of address critical supply/demand realities fully. They’re all just getting in the way of economic growth.

  • Roslyn Layton

    Dear Michael,
    Thank you for taking the time to read and respond. Please know that I value your input.

    In response to your comment, I give you the conclusion of the academic literature of on unbundling (what you call open access) starting with Crawford herself
    1. Susan P. Crawford: “Mandated unbundling under the Act is widely viewed to have been a failure.” Network Rules, Law & Contemporary Problems, Spring 2007, at 51, n.16
    2. Richard A. Epstein: “There is widespread agreement today on all sides of the telecommunications wars that something is deeply flawed with the design or implementation (or both) of the Telecommunications Act of 1996.” Takings, Commons and Associations: Why the Telecommunications Act of 1996 Misfired. 22 Yale Journal on Regulation 315, 315-6 (2005).
    3. Mark A. Lemley & Philip J. Weiser: “The unbundling regime of the 1996 Act represents, on almost all accounts, a policy failure.” Should Property or Liability Rules Govern Information? 85 Texas Law Review. 783, 812 (2007)
    4. Kevin Werbach: “The current legal framework, embodied in the Telecommunications Act of 1996 is widely regarded as a colossal failure.” Only Connect, 22 Berkeley Technology & Law Journal. 1233, 1237 (2007).
    5. Robert W. Crandall & Leonard Waverman: Interestingly, competition for local telephone services has arisen intermodally, with wireless telephony now serving as an effective substitute for traditional wireline services. Interestingly, regulators were initially reluctant to regard cellular telephones as substitutes for wireline service. The Failure of Competitive Entry into Fixed-Line Telecommunications: Who Is at Fault?, 2 Journal of Competition & Economics 113, 114-25 (2006).

  • Michael Elling

    1) yes, all “competitive apologists” or vertical monopoly supporters (the latter divided between private or government controlled/sponsored at layer 1)
    2) if you go to my website under archives/regulatory, you’ll see that I already referred to the Telecom Act as a well intentioned “farce” back in 1996-97
    3) I did this because the Act lacked teeth and it was not applied multilaterally and across the board and was therefore arbitrary and capricious, which is how the Bells defeated many measures
    4) I was a bear on the $250bn CLEC binge because of their adherance to a vertical service model and the 3 layer “IP this and IP that” communications miracle touted by the likes of Nacchio, Grubman and Ebbers; all, unfortunate contemporaries
    5) At the same time I roundly criticized Telric for being not forward looking, but rather backward looking based on average costs. estimating marginal cost a priori requires numerous supply/demand iterations which I describe on my blog. in addition i relate how:
    6) in 1996 I published a seminal piece on wireless substitution called the 4Cs (cost, coverage, capacity, clarity) based on the revolutionary 10 cent digital model which accurately forecast where we are today
    7) i’ve since published the law of wireless gravity which holds a wireless bit will seek out fiber as quickly and cheaply as possible. as well you can read about my debunking of “intermodal” competition
    8) i don’t make stuff up. i analyze real world pricing models, focusing on 8 major quantitatively and qualitatively scored supply and demand drivers.
    9) this is how i explain/understand things like WAN-side scale, how Google is able to price a Gpbs and how 802.11 has performed so well.
    It’s time for a major rethink to policy driven on the real-world examples of what worked in the 1980s-90s; not what didn’t work and was so evident at the time.

  • Richard Bennett

    It’s wildly inaccurate to claim that Crawford, Werbach, Lemley, and Weiser are either “competitive apologists” or vertical monopoly supporters. I can’t comment on the others as I’m less familiar with their work, but your comment is simply false.

  • Michael Elling

    As Roslyn points out, they believe competition failed. And they don’t understand fully why it failed, instead relying on the false notion of natural monopoly. Nor do they offer any competitive approaches/models that might succeed, instead supporting a government interference/oversight/overbuild model, which is socialism, which is monopoly.

    The ONLY potential for monopoly is the ROW or frequency (as with any property), and that’s an unnatural one in the case of communication networks which are eminently shareable. Why do you think AT&T put in a 50 mile limit in the Kingsbury Commitment 100 years ago? Because radio was already a potential competitive bypass solution in the mid and last mile at that time. Imagine where we would be today if we had taken the “interconnect” fork in the road back then.

    Only government barriers/restrictions sustain information monopolies; and they can and do occur at many different layers of the InfoStack; copyright, address directories, settlements, operating systems, cartel standards, etc…; not just layers 1-2. (The InfoStack is a free model to objectively analyze and discuss all the parameters in this debate.)

    There is nothing in network theory which substantiates or supports monopoly in the short or long run, particularly today where supply and demand are so dynamic and the range of outcomes infinite. In fact just the opposite. The competitive WAN, wireless, data/internet and 802.11 examples over the past 30 years have proven the case for horizontal, competitive, digitally priced, generative supply/demand ecosystems.

    In these horizontal ecosystems, pricing reflects marginal cost at every layer and boundary point “a priori” to generate enormous demand elasticities. This approach contrasts with average pricing ex poste by inefficient vertical monopoly service providers that retards demand. We see a lot of that around now due to remonopolization in layers 1-2. In the horizontal model pricing drops 99% over 10 years due to the combination of effects from moore’s and metcalfe’s laws, yet overall revenues grow due to price, private back to public, and 3rd party application elasticies. The model only fails when it hits government imposed barriers. This is the answer the competitive apologists are missing as to what happened over the past 30 years.

  • Richard Bennett

    Spectrum isn’t as shareable as many people seem to think. Each Wi-Fi access point can only handle one packet at a time.

  • Michael Elling

    Sure Richard. Add this to the amazingingly bizarre, incongruous and disingenuous arguments you make that simply don’t make any sense.

  • Richard Bennett

    Please explain what you think is incorrect about my statement.

    Clue: The WiFi Medium Access Control protocol is called “Carrier Sense Multiple Access with Collision Avoidance”(CSMA/CA). Why is there a need to avoid collisions?

  • Michael Elling

    Richard, give it up. SMDS lost the battle about 22 years ago to frame relay and then to IP. Ethernet became the dominant LAN protocol and scaled into Wifi and the MAN. Now it’s eating the WAN in layer 2. What’s your point about congestion?

    Verizon just announced 200 gbps in one wavelength between NYC-Boston. That’s 88 wavelengths per fiber pair. Assume 196 pairs…that’s a lot of bandwidth.

    What is your point about congestion? That’s the whole concept of packet systems. It’s why the stupid AT&T executives laughed at the IP guys in the 1970s. They didn’t get it.

    What is your point? Other than revealing that you hold to 30+ year old conventions? And secretly wish AT&T had never been broken up?

  • Michael Elling

    Richard, it’s pretty laughable that you question the scale and commercial success of Wifi due to it’s inherent “sharability”. It really points to the flaw in your understanding of networks and service provider business models.

  • Richard Bennett

    Excuse me? I simply said that Wi-Fi access points can only receive one packet at a time (per frequency band if you have a simultaneous dual band access point.)

    That’s a technical fact, like it or not, that relates to the nature of wireless networks and has nothing to do with anybody’s business models. If you think it’s wrong, make a pertinent argument.

  • Richard Bennett

    That’s a completely irrelevant piece of trivia, but it’s funny that you think you need to lecture me about how “Ethernet scaled into Wi-Fi.” If you only knew.

  • Michael Elling

    “That’s a technical fact that relates to the nature of wireless networks”. So now you’re “one packet at a time” refers to all wireless?

  • Michael Elling

    I guess National Instruments knows. “Wi-Fi is an extension of the LAN to the wireless domain (WLAN). Wi-Fi technology has simplified the installation and distribution of networking infrastructure by replacing wire cabling with low-power radio waves. Because it is standards-based, it is widely available and integrates seamlessly with existing Ethernet networks. For wireless data acquisition, Wi-Fi is an easy way to tie into existing corporate infrastructure without special gateways or converters.

    Furthermore, it not only scaled in corporate LAN environments, but cable modem environments for residential ubiquity, and once Intel launched Centrino in 2003, Wifi pretty much dominated the LAN/PAN boundary point for access and today it accounts for nearly 40% of TCP/IP access vs 2-3% for mobile networks. All market driven. No govt dictates for 802.11 other than assigning the bands and power limits. (The latter providing near infinite reuse (cellularization) of spectrum.) That’s pretty good free-market “sharing” and efficiency in my opinion.

    I provided you with several pieces of “trivia” about scaling and dominance of protocol stacks that relate to WLAN scaling and sharability,not just one, so that you could see a pattern. It’s good to know you are a dyed in the wool monopolist as you didn’t counter the point about the break-up of MaBell and what king of service-stack hell we would be in today. Probably some Orwellian future or something resembling the infrastructure and networks in the 1985 Gilliam/Pryce/DeNiro classic, Brazil. Funny that, 2 years after the break up. What they didn’t know!

  • Richard Bennett

    I was one of the people who designed hub-and-spoke Ethernet in 1984 and also one of those who designed Wi-Fi in 1990. It’s great that you’re enthusiastic about these inventions, but a little sad that you don’t seem to understand them very well. Such is life.

  • Michael Elling

    Richard, you’re one with the rebukes, rarely with civility. Too bad you can’t take your knowledge and pedigree and apply it to a competitive, generative business model.

  • Richard Bennett

    That’s actually what I’m doing; socialism hasn’t been my thing since 1969.

  • Michael Elling

    Monopoly is present in extreme forms of socialism and capitalism. So its not really a political thing.

  • Pingback: dortgurtgiretl

  • Pingback: paleo cookbook recipes

  • Pingback: Cumbrian Plumbers

  • Pingback: 3 day notice california

  • Pingback: Google

  • Pingback: New York City Tax Fraud Attorney

  • Pingback: Houston movers

  • Pingback: water ionizer scam

  • Pingback: Trade Show Exhibit Displays

  • Pingback: insuranc rates by zip code

  • Pingback: see this

  • Pingback: fifty shades movie

Previous post:

Next post: