Duopoly Carriers Aren’t Evil. They’re Public Policy.

Susan Crawford at Code for America

Susan Crawford at Code for America

Last week I attended a presentation by Susan Crawford, the author of Captive Audience, The Telecom Industry and Monopoly Power in the New Gilded Age. It was held, fittingly, at Code for America in San Francisco.

Ms. Crawford has a pragmatic outlook on the state of telecommunications in America, and I think she’s right on. In highlighting the outcomes of our largely duopoly telecom marketplace in America, in a recent NY Times article, she says “It isn’t evil, it’s just the way things work.”

The fact is, it’s not just the way things work, it is the way the appointed regulators at the FCC from 2001 to 2008 intended it to work. The goal during the chairmanships of Michael Powell and Kevin Martin was “robust, intermodal competition”, where Big Cable and Big Telco would complete, without the distraction of additional competition. In other words, duopoly. And, it’s what we got, by and large. (More on this later; Sonic is the exception.)

So, it’s not evil on the part of the two firms who dominate each regional market, it’s simply the outcome of our regulatory policy.

FCC Chairmain Powell’s BrandX and TRRO, and Chairman Martin’s Broadband Forbearance order together largely eliminated the potential for uniform nationwide competitive pressure on the two operators in each region. The goal was to limit competition in hopes of spurring investment by these two incumbents, with Martin saying about the Forbearance order:

“Promoting broadband deployment is one of the highest priorities of the FCC. To accomplish this goal, the Commission seeks to establish a policy environment that facilitates and encourages broadband investment, allowing market forces to deliver the benefits of broadband to consumers.” -FCC Chairman Kevin Martin

In other words, handing over a doupoly is supposed to encourage the two appointed competitors to duke it out.

I will concede that the policy has worked to some degree. Cable firms have rolled out DOCSIS 3.0 upgrades, and both AT&T and Verizon have made investments to catch up in video and increase broadband speeds. Even Centurylink has made some upgrades to their network.

In a recent article, Susan Crawford points out an op-ed by a Cable executive, titled “U.S. the leader on broadband“. Maybe by making it the headline, it becomes truth.

But, the fact is that we remain well below even the OECD median speed, and we are certainly no leader like Japan, Sweden or Norway. I suppose we can at least take solace that the offered broadband speeds in the US are not as bad as Mexico or Greece.

OECD Broadband Speeds Chart

OECD Broadband Speeds Chart

The policies which arguably helped drive these investments have come at a significant cost to consumers — Americans pay far more for their broadband and get less than citizens of most other developed nations. We pay more than three times as much per megabit as citizens of Sweden or Japan, and a whopping sixteen times as much as those in Korea. In fact, Americans pay more for broadband than residents of virtually all of the OECD countries. Only five countries do worse than US residents: Turkey, Chile, Poland, Greece and Mexico. We are in sad company when it comes to what we pay for broadband.

But, that’s not evil, that’s the outcome of our policy. Cost is the casualty when an intermodal duopoly is favored over full competition as a public policy goal.

OECD Broadband Prices Chart

OECD Broadband Prices Chart

The US median price per megabit is $5.42. Did we get enough performance in exchange for the creation of a virtual duopoly? Probably not, but that’s not evil on the part of our dominant broadband providers, that’s policy, and the way the marketplace was intended to work.

Fusion logo


But what about Sonic.net’s Fusion service?

Fusion is the sort of competition that the 1996 Telecom Act envisioned. And, we are still here despite the de-regulatory hit job by Powell and Martin. We are still here because the basic bones of the 1996 Telecom Act remain strong. While competitive access was curtailed by the intermodal duopoly agenda of the Bush-era FCC, competition is still possible under the Act, at least in metropolitan regions.

Sonic.net’s Fusion service offers competitive European style pricing and features, an “all-in-one” service delivering uncapped and unlimited broadband at up to 20Mbps, plus unlimited nationwide phone service together for one low price. When both the voice and data products are considered, it offers an advertised price about $1 per megabit, better than all but a few of the OECD nations, and less than one fifth of the US median price.

Fusion is available today to four million homes in ninety-six California cities around Los Angeles, the Bay Area and Sacramento. Join up. Then tell a friend. It’s a growing revolution.

  • Maenad

    Please bring service to Carmel Valley, CA. I want my Sonic.net back. I’m stuck with AT&T out here, and using local Redshift on their copper.

  • Parker

    What’s interesting is in the United States, you would think just because you live in a city where there is a decent population density, that you would have more options available (or at least fast speeds available). Cities like San Francisco prove that to be wrong. While there are a couple of reasonably priced options (hi Sonic!), there aren’t a lot of reasonably priced 50+ Mbps options. I lived in Madrid and could get 100Mbps for about $45USD/month. Most of the large cities in Spain could get something similar, granted my evidence is anecdotal. Having said that, we do have a much more stable economy than Spain, it’s just interesting to see that in their cities they do have better options than most of us in the Bay Area.

    Maybe the Bay Area is a unique case. I know Sonic would have started deploying Fiber already in San Francisco had it not been from slow moving political processes. However, other US cities don’t seem to be faring much better.

  • plwww

    I love sonic.net, but I don’t believe using “advertised speeds” is at all useful for actual comparisons(especially when criticizing policy, though it is indeed flawed). When I see real-world speed data, the situation looks far better for the U.S; we still aren’t “leading” but we’re usually above the average, and the top-end isn’t too far out of reach. For example: http://www.akamai.com/stateoftheinternet/ : On the map for “Average connection speed” Japan, leading in the “advertised speed” chart with ~150mb, instead comes in with 10.8mb while the U.S. gets a respectable 7.6mb. Sweden, second place on the “advertised speed” chart with ~101mb, comes in behind the U.S in the Akamai chart at ~7.2mb. Obviously, a single company’s real-world data is also an incomplete picture of broadband speeds, but seems much more credible than the “advertised” speeds used in the post, and creates a much different picture.

  • Realist

    I wish that the general population would understand that sonic.net is not even competition for AT&T. Back in 1996 when the telecom deregulation happened the ILEC’s AT&T and Verizon etc… We’re forced to lease lines to CLEC’s sonic.net etc. so the fact is that sonic.net is more of a AT&T reseller then anything else. I have found NO other agreements between sonic.net and any other ILEC’s plus sonic.net has almost no control over the way the lines are connected etc. they may be who the end user talks to and pays but in all reality it is still AT&T service. People need to understand that sonic.net is not competition.

  • sonicnet

    This claim is a bit misleading. Sonic.net (and our utility subsidiary Sonic Telecom) is not a reseller of AT&T services – we are a full facilities based competitive local exchange carrier. This means it is our equipment in the telco central office building that customers are connected to, and our backhaul network, our IP transit, etc.

    It’s may be a bit confusing because we used to primarily serve customers using AT&T’s DSL. That service was their equipment (DSLAMs) in the central offices, and it was their backhaul network. The customers were then handed off to us on large aggregation circuits at layer 2 (ATM), and we would do the Internet connectivity, services, sales and support. But it was, strictly speaking, resale a DSL line.

    As a full facilities based carrier, the Fusion service is our own DSLAM (which uses newer technology than their old ADSL1 network, ours supports 24Mbps maximum instead of 8Mbps, and bonding for double that), our own voice service and switching, the backhaul network, and then as before the other portions – Internet, services, etc. We only rent the copper pair itself from AT&T – all of the equipment that connect to BOTH ends of that copper is all Sonic.


  • Jay Gamel

    If you could only bring it to Kenwood. We do have optical cable to here, I understand.

  • fredfnord

    Akamai statistics are heavily skewed by corporate networks, which is why people don’t really use them to measure things like home networking speeds. One corporation with a 150 mbps line (as mine has) and 200 users who use the internet from there can skew the statistics of an entire town to complete unreliability. (200 users at 150 mbps plus 10,000 users at 2 mbps = an average speed of 5 mbps.) The best numbers we have on the US are found elsewhere, and they are pretty pathetic.

    It does make me wonder about Japan, though.

  • fredfnord

    You’re lucky. I live in SF too, and the best I can get without rewiring my apartment for cable modem at my own expense (plus that would mean dealing with comcast, yuck) is Fusion dual-bonded-pair at 4 mbps per line, for a grand total of, in practice, about 6.8mbps.

  • davidahn

    I know it’s already been a few months since this post, but a note on statistics: median values are more representative of reality than averages, as medians are less skewed by outliers. Looking at median bandwidth, Korea seems to have BY FAR the highest median bandwidth.

    I’d love to see the US take over leadership (or at least get close), but the cable companies and Verizon seem to have agreed to a stalemate to reduce infrastructure expenditures and maximizing profits at the expense of end users. I’m “lucky” to have Verizon’s 85/35 service for $89.99 in our Apple Valley, CA house, but in our San Diego house, Cox offers 50/6 for $99.99 (and that’s the fastest widely available speed in SD (Webpass supplies urban buildings in San Diego with 200 Mbps service for $50/mo, but I don’t want to move and pay $900/mo in HOAs just for fast internet). Pathetic when Sonic, Google, and others offer 1 Gbps for $69.95.

    Dane, can you point me in the right direction so I can establish a local initiative to get gigabit internet here in beautiful Bonita, California?

  • sonicnet


    Clearly, we share your concern about US broadband and competition, and we’re working to improve the situation. I don’t have any specific pointers for achieving Fiber-to-the-premise service in your specific location. Clearly it can be done by a determined entity and at some cost, as we’ve shown.

  • davidahn

    Thanks for all you do for your customers. Wish we had an ISP that cares about more than profits down in SD (outside of downtown, that is)!

  • So, who takes care of the wires (until fiber gets installed?)

    My new house (remodel, legally) is almost done.

    I have Sonic there now.

    We have underground utilities.

    PGE replaced the electrical service from the transformer to our building (at huge expense.)

    The contractor wants to replace the phone wires as well.

    Does he call ATT? Sonic?

    Anything special we can do to get ready for Sonic fiber-to-the-home?



  • I’m in San Francisco (user “baggins”)