A Model For A Competitive Broadband Market

In the White House’s Broadband Report, released yesterday, it states:

At the same time, limited competition is also a challenge even in communities with high rates of adoption. Today, nearly 40 percent of American households either do not have the option of purchasing a wired 10 Mbps connection or they must buy it from a single provider. Three out of four Americans do not have a choice of providers for broadband at 25 Mbps, the speed increasingly recognized as a baseline for broadband access. Lowering barriers to deployment and fostering market competition can drive down price, increase speeds, and improve service and adoption rates across all markets. 

That last bit, about competition driving down prices and increasing the quality of broadband services, is exactly right. But how do we get a competitive broadband market?

I believe the telecommunications market needs to move away from vertical integration where one provider builds, manages, and delivers the entire telecommunications stack to the market. We need to move to a layered model much like the software industry has adopted and which results in a highly competitive market at each layer. We tried this in the 1990s with the Telecommunications Act of 1996 which provided that incumbent carriers were required to offer new entrants access to their networks. For many reasons, that didn’t work out and tens of billions (maybe hundreds) were incinerated when the CLECs (the new entrants) were stymied by the incumbent carriers and failed to build sustainable businesses. That massive failure has weighed over the broadband market for the past 15 years as most investors moved onto more fertile areas of investing in the TMT (telecom, media, technology) sector.

But just because it didn’t work doesn’t mean it was not the right model. The owners of the plumbing (copper, fiber, spectrum) must be required and incentivized to open their networks to competition. When this happens, we see entrepreneurs emerge to offer new innovative services and investment dollars follow them. When this does not happen, we see wasting assets and stagnating performance and quality.

How do we get there? Through a combination of smart and lightweight regulation (oh the horror of that word!) and by focusing on and supporting innovative entrepreneurs who are working in the telecommunications market. We are doing both at USV. We have been steady, vocal, and critical supporters of regulatory efforts, championing them when they are light and smart and opposing them when they are heavy and dumb. We are also making investments in new innovative telecommunications services. We have announced at least one of them and have at least one that we have yet to announce. My partner Brad is leading our effort in this area both on the policy side and the investment side.

I believe in competition more than any other market or governmental force to bring good things to the market. We need more of it, not less of it, in the broadband market. If we want to expand broadband access to every home, business, and person in this country, competition is the way to do it.


Comments (Archived):

  1. Twain Twain

    The telco market is clearly inefficient as can be seen by falling transit prices.The sooner incumbent carriers open up to innovation, the sooner we get mesh networks delivered by individuals via IoT devices no bigger than a button whereby it authenticates, provisions for privacy and each individual can monetize their own data and bandwidth.Yes, this “crazy” idea is entirely possible because the technical pieces (chipboard sensor, wireless battery charging, software, OAUTH and Blockchain) are emerging and there are smart people who’d know how to harness and integrate these technologies to deliver much more efficient and monetizable data carriage and analytics.

    1. Twain Twain

      Also adding this because it provides a clue to the layers that future models would include. Now, the thing is we’ve moved from copper wires to fiber optics. There’s research in materials science of a new material that can be both an insulator and a super-conductor — which would be transformative if we can figure out how to stabilize the two states (it’s all to do with Quantum super-position).In 10 years’ time, we may discover we don’t even need copper wires or fiber optics cables because a team of scientists has discovered something even more amazing than frequency hopping (Hedy Lamar and George Antheil, by the way).

      1. Twain Twain

        Do we NEED radio masts that size?Remember the first computers? They were the size of a room. Now, our smartphones have the same power in a fraction of the size.So we don’t need radio masts that size and we can get them down to smaller form factors — maybe size of a paperweight.It’s funny because just this morning whilst walking, I was thinking about telco radio waves; specifically in relation to RFID tags and all that copper.It makes 0 sense to me that the copper gets coiled in this way.So….I’ve made a new design for my IoT device.

        1. Michael Elling

          The law of wireless gravity holds that “a wireless bit will seek out fiber as quickly and efficiently as possible.” I love wireless, but I fully appreciate its limitations. Again, it’s another type of layer 1-3 tradeoff that I mention above.

          1. Twain Twain

            You do know I’m in a personal war with bits, right? Haha.Let’s imagine there are no bits, no 1’s and 0’s. What would computing look like and do then?I proviso this with John Von Neumann’s comment: “When we talk mathematics, we may be discussing a secondary language built on the primary language of the nervous system.”If mathematics is a secondary language AND it’s the slow System 2 that Daniel Kahneman refers to in ‘Thinking Fast and Slow’, then how much faster and more powerful would a computer, telecoms network and code that was more based on the carbon and oxygen matter of our nervous system be — rather than based on mathematical bits?

          2. Stephen Voris

            I have to wonder what humans would look like if there were no bits – if “computing” had no absolutes. I suspect we didn’t make computers because we were already good at what they do, after all – System 2 is, as you’ve noted, slow. But then, that’s System 1 thinking from me.The concept of System 2 versus System 1 thinking is (in my mind, anyway) a tradeoff between accuracy and speed (or depth and breadth) – between rules, and rules of thumb.Faster… unlikely, actually. Nerve impulses themselves only get up to ~300mph, though we can suppose this would be optimized in our hypothetical bitless computers in the same way silicon chips are being optimized in bit-based computers.Worse, though, would be the loss in accuracy. If you’ve ever played “telephone” – put a group of kids in a circle, whisper a phrase in one’s ear, have them whisper what they hear to the next kid, and so on – you know that what comes out frequently has only a passing resemblance to what went in.You can see the same sort of behavior with, say, Google Translate and multiple languages.I suspect that wasn’t exactly what you meant, but then, that too sort of illustrates my point.On a bit of a tangent, though, I do seem to remember reading about someone working on computers that sacrifice some accuracy for efficiency; this was a few years ago, though, and I can’t remember the source.

          3. Twain Twain

            A few observations:(1.) Humans aren’t made of bits, 1’s and 0’s. We’re made of atoms of C, H, O, N etc.(2.) It’s because computers currently have absolutes (0=negative, 1 = positive) that they’re functionally incapable of processing ambiguities.Ergo, AI problems which Yann Le Cun, AI guru of Facebook, calls “one of the biggest, most complicated scientific challenges of our time” — principally, humanity’s collective knowledge of how to better emulate intelligence with machinery and Natural Language understanding.(3.) Accuracy and efficiency need not be each other’s tradeoff. Just as correlation and causation can co-exist in a piece of data.It’s simply a matter of the inventor(s) solving the AI problems differently.

          4. Stephen Voris

            1. By that standard, computers aren’t made of bits either – they’re made of atoms of Si, usually. And if you look at what humans do at the neuron level, it looks an awful lot like bits – at any given time neurons are either “firing” or they aren’t. Bits might not be the whole story – the circumstances under which any particular neuron decides to fire are frequently a lot more complex than any single logic gate, after all – but so far as we can tell right now they’re a reasonable shorthand for how neurons communicate with each other.2. Disagree, here; presence of absolutes and absence of ambiguities are distinct concepts. Color, for example, can be an ambiguous concept, but that hasn’t stopped computers from being able to produce them (or ‘bit strings representing them’ if we want to be pedantic).3. Maybe not, but that tradeoff seems to show up an awful lot whenever it comes to saying something versus proving it. Error correction has a cost, and it’s one that can only be reduced, not eliminated.

    2. Wyatt Brown

      Thank you for charts. We love charts! 🙂

    3. Michael Elling

      Mesh is an over-used term. In all networks there is a tradeoff between layers 1-3 in terms of reducing cost and addressing capacity issues. Another way of saying is that change occurs at the edge and collapses to the core (aka a switch) or vice versa. What’s missing, as I said above, is true (inter)network effect (Metcalf’s law in addition to Moore’s) brought on by the right type of settlement systems between the actors. Bill and keep fosters monopolies, which is presently where we are at the core and edge.

      1. Twain Twain

        I agree about the over-use of “mesh networks”. Personally, I’d prefer it to be coherency spheres but we work with inherited lexicon…Re. the layers I see that structure in Convolutional Neural Networks.Yes, Metcalf and Moore’s gets quoted a lot there too.However, there’s also some Hopfield which affects how efficiently the data transfer happens between layers and the cycle of the collapse.There’s still invention needed in the switch and this is why carbon nanotubes are a fascinating development:* http://spectrum.ieee.org/na

  2. jason wright

    https://en.wikipedia.org/wi…isn’t there a satellite internet project in the works that would end this inequity, and not just for Americans, but also the entire global population?

    1. sigmaalgebra

      Maybe satellites could work well enough for streaming, say, movies. But for the usual interactive communications needs of phones and computers, the speed of light is too slow, and there is too much delay, 1/2 second or so (computer to satellite to ground station to server to ground station to satellite to computer, four trips of about 22,000 miles each plus the rest, and that’s nearly a full second).

      1. Rob Larson

        The 22,000 miles you mention is for geosynchronous satellites. Elon Musk is planning (as is Greg Wyler @ OneWeb) to deploy satellites in low earth orbit (<775 miles). That brings travel time to 0.004 seconds for one leg, or 0.016 seconds for all 4 legs. Which is comparable to fiber optic or cable broadband latency.According to FCC, the average latency for broadband service last year was 0.035 seconds.

        1. sigmaalgebra

          Musk, hmm?> geosynchronousThat’s how the heck the computer knows where to aim its antenna, both to send it signal to the satellite and how to receive the relatively weak signal from the satellite. Else we’re talking a lot more power for the transmitters at both ends.I know: Some years ago there was an effort with a swarm of low earth orbit satellites for mobile voice telephone. Apparently that’s why now we have cell phone towers!And there’s Page/Brin with blimps, but that maybe only for, say, some rural parts of Africa or some such.But, net, I have to suspect that mostly Fred’s talking about the last mile.For the last mile, there are more approaches: Can have some phased array beam steering with a lot of fast Fourier transform arithmetic, etc. to get more data rate out of limited frequency bandwidth.As I recall, @JLM has multiple Gbe last mile sources — ask him how that was done. Maybe his ‘community’ put in the last mile physical connections and routers, and then metro providers get to connect.

          1. Rob Larson

            Musk is planning to get around the “where to aim the antenna” issue by blanketing the earth with a network of 4000 small satellites. (compared to ~1000 total satellites currently in operation).This will include the last mile. Pizza-box sized receiver on the roof, which then beams signal to computers at speeds “much faster” than what is currently available via fiber.

          2. sigmaalgebra

            The pizza box “receiver” and its “beams” are for the neighborhood? If so, then why have the pizza box communicate with satellites instead of just something on the ground? Installing coax in the ground, etc. is actually not so expensive anymore, and if one pizza box on one roof can serve, say, a neighborhood, then that’s just one coax per neighborhood and likely much cheaper than satellites.The pizza box has an array of many dipole antennas and does phased array steering to aim the array sensitivity to reduce total radiated power for both the pizza box and the satellites?Ah, I worked too hard with the fast Fourier transform for acoustic signals and arrays!

          3. Rob Larson

            The purpose of the pizza box is to communicate with satellites. The insides haven’t been disclosed, but your idea of dipole antenna array is interesting. Then again I studied physics not engineering so I’m not sure what’s practical in the real world.Then it can either beam wifi to the computers in the house or you run coaxial to however many computers/houses you want to share with. Though I believe it will be cheaper to buy another pizza box than run cable underground.The interesting thing is that the fewer people accessing a given satellite at a time, the less bandwidth has to be shared – therefore you will get the fastest speeds in the most remote areas.

          4. sigmaalgebra

            Okay, have a pizza box per suburban house.For antenna patterns, it’s just some Fourier theory and basic E&M. You can look it up and read it in 10 minutes or derive it yourself in 15 minutes. Of course, as is curious, the math for the pattern for both a receiver and a transmitter are the same.The US military is all over such beam forming via phased arrays, for both radar and sonar. So, for a receiver, take the signals from all the dipoles in an array and, to track 20 targets, add up the signals 20 different times, each time to steer a beam of antenna sensitivity at one of the 20 targets. E.g., take the nose cone off the front of a high end US fighter plane and will likely see a pizza pan of dipoles. and that’s the radar antenna. One big application was to Aegis for the US Navy: The whole front of the ship was some big, flat panel, and under it was, likely, a lot of dipoles. Another application was for ABM radar: A building with a big, flat side, and under that a lot of dipoles.

  3. Avram

    Why does smart has to go with light in terms of regulation? Can’t it be just smart vs dumb regulation? Sometimes light regulation can be dumber than heavy regulation!

    1. jason wright

      light regulation gives people the room to enact their smartness?heavy suppresses opportunity?there has to be an incentive to motivate people to action.

      1. Avram

        I agree we should always move for simpler but the objective should be effectively regulating not simple for the sake of simple.

        1. jason wright

          true. start with a little reg and if it proves to be not enough then add a little more, with a tweak here and there. real reg for real issues. i hate doctrines.

    2. fredwilson

      i believe that the smartest regulation is light touch regulation. so i conflate the two. but i take your point

      1. Jamie Friedrech

        Has there ever been ‘light regulation’? Where once the objectives are achieved, those working for the program will gladly leave their own jobs and find another problem to solve?

        1. pointsnfigures

          Once regulation begins, lobbying begins. See Professor George Stigler’s studies. As a matter of fact, big huge corporations benefit from heavier regulation. They can afford to lobby for loopholes, and lobby for ways to erect regulatory barriers to competition.For example, suppose you wanted to start a Bitcoin Futures Exchange and compete. Perhaps you want to list some other products where you see an opportunity. Currently, there are 2 dominant futures exchanges in the US that trade everything (oligopoly) To startup, before you have even traded one contract, hired anyone etc, the tab is $2M minimum. $1.4M of that tab is due to heavy regulatory barriers set up by the CFTC and Dodd-Frank.We see this in the oil industry, health care, banking, telecom, and so on and so on.

          1. Avram

            More money having a bigger say in the gov’t is the problem, not regulations.

          2. pointsnfigures

            http://www.sjsu.edu/faculty…, the more regulation, the more that big money enters. It’s a cycle. To break the cycle, deregulate, and increase competition. The more competition gets regulated, the less competition there is.

        2. Avram

          There’s constant evolution in the industry – so regulators will have to make sure regulation evolves alongside.

          1. pointsnfigures

            Market is always faster. Government can never keep up.

          2. Michael Elling

            Governmental actors don’t take risk. Therefore they do not “learn” how to allocate capital efficiently. They also govern to the average, not the marginal segments, whereas markets clear marginal supply and demand better.

      2. Avram

        Regulations are the tool to make sure the otherwise free market considers externalities, i.e. carbon emissions has negative effects that currently can’t be priced by the free market so the gov’t regulates them. We should always find simpler ways but without losing sight of the objective.

        1. pointsnfigures

          To a point. sure, there are negative externalities. But, the government needs to tread very carefully when it regulates them. Even carbon might be better off regulated by a network, not a government agency.

        2. Stephen Voris

          Everything has externalities, and that includes regulations.One of the negative externalities of regulations (or laws, or processes) is what programmers call “technical debt” – the amount of time someone new to the field (or society, or job) has to spend reading the manual and understanding what it all means before they can really get started.In other words, all regulation inherently favors the incumbent(s), because they can start paying down that technical debt immediately. That bias may be worth the benefit the regulation provides, but it’s hardly guaranteed.

    3. JamesHRH

      Heavy regulation is bad regulation.Simple principles, precise standards and clear penalties.Most regulations look like fungus, growing in asymmetrical patterns based on a path of least resistance.

  4. creative group

    One key point in establishing competition in the broadband space is legislation. That means buying legislators. The lobby that keeps this monopoly going is using billions to maintain the poor service each community complains about.

  5. David Morken

    We operate the last CLEC ever built in the US. The crux is how to achieve this: “The owners of the plumbing (copper, fiber, spectrum) must be required and incentivized to open their networks to competition.” The 1996 act was 750,000 words long and was considered heavy, but wasn’t heavy enough even with hundreds of billions behind it. Are you implying a novel new lightweight approach or something like ObamaCom to go with ObamaCare? Or an imminent domain tack like the FCC repacking spectrum over broadcaster objections? Or…

    1. Peter Radizeski

      the last CLEC?!

    2. Michael Elling

      TA96 was a well-intentioned farce that didn’t have teeth and set the wrong set of actors playing against eachother. TELRIC was a disaster as govt cannot set pricing efficiently. Pricing is only arrived at efficiently by clearing supply and demand iteratively ex ante at the margin. Govts (and monopolies) only understand averages and are terrible future predictors. There are many things that we got right from the late 1970s to the mid 1990s. But once we had TA96 and wireless auctions that didn’t mandate interconnection (at the handset) we jumped off the competitive rails in the late 1990s. It was only Steve Jobs forcing equal access (aka wifi offload) on AT&T in 2006-07 (not sure exactly when the deal was struck) that competition has finally and partially been resurrected. And with the current IP stack model, that’s a duopoly, not really a competitive market. And worse, look at the concentration of advertising in mobile.

  6. Oscar Jung

    In Denmark where I grew up there is much better competition as the owners of the infrastructure and licenses (3G, 4G etc) are forced to rent out the infrastructure to smaller carriers. In general there is a move to decouple the different layers of the business. We see the same in energy where the larger energy companies are forced to break up so it is not the same company that owns the power station, the grid, the maintenance and the sales. When I moved to US I was surprised that data and cell phone subscriptions are so expensive and that coverage is so bad…

  7. andyswan

    I will spare you my “role of federal govt” and “unintended consequences this is hilarious” speech :)Instead, I will do what I do best–predict the future:The goal will be achieved. It won’t come through regulation (light or heavy). It won’t rise from the burial ground of past endeveours.It will come out of the clear blue sky. We will think of it as a gift from above. The source will be a company that went where the light-fingered regulators are least powerful.It will transform humanity. It will create wealth across the globe as never seen before. Its creator will be villivied for the wealth he achieves. Eventually the service itself will be taxed and regulated into obscurity by the envious masses.And then another will rise.

    1. jason wright

      “…he achieves.”? perhaps i can see why at least half of the masses might be envious.tax can be a good thing. go to Denmark for your next holiday.

      1. pointsnfigures

        No thanks. Remember when Oprah went to Denmark and asked “Where’s all your stuff?” and the people said, “We don’t have any stuff because we can’t afford any stuff.”

        1. Anne Libby

          Yet here in NYC, people in all of my communities are talking about “konmari”-ing All Of The Things.https://en.wikipedia.org/wi…Because many of us have too much freaking stuff.

          1. pointsnfigures

            The point is, there is no economic opportunity in Denmark to even begin to hope for affording all that stuff. I’d rather have the problem of too much stuff than not having the opportunity to acquire it in the first place. Heavy regulation does that to a society, and socialization of resources does that to a society. It’s why they came over here.

          2. Anne Libby

            Heh, in a rush out the door, I see that sometimes, my tongue-in-cheek doesn’t come across online. (again, sigh.)I hope someone from Denmark pipes up here. My general impression of that part of the world is prosperity.

          3. pointsnfigures

            My friend Mike Gibbs, a labor economist at Chicago Booth, has done a lot of economic research in Denmark. The one big difference between there and the US is population diversity and size.

          4. andyswan

            Isn’t this where we start chanting about diversity being a strength

          5. Twain Twain

            Diversity is a strength. Some of the top companies have been founded by people from all over the world who emigrated to the US.Google being an example.

        2. jason wright

          consistently rated as the country with the highest quality of life and overall happiness. of course without enough stuff those things count for little.

          1. Drew Meyers


        3. Drew Meyers

          “stuff” doesn’t provide people fulfillment in life. I’m not sure they are missing out that much.

    2. sigmaalgebra

      I got it! I figured out what you have in mind! You have us using the Star Trek/Star Wars subspace radios!In the meanwhile somehow I’m suspecting that we are talking about a lot of “heavy” regulation of the physically connected (twisted pair copper, coaxial, fiber) last mile, especially for residential suburbs.But, maybe for the residential suburbs, to get around digging up the streets and yards, in each neighborhood have a pole in the ground with a wireless node at the top that communicates with antennas at each of the houses.And why don’t we have that yet? Maybe because no one sees much chance of significant profit in it.Why? Because the market is not monolithic, that is, has split into fragments. E.g., there are a lot of mobile wireless devices out there, and they have been taking away much of the reason for the physical connections. So, I’m guessing, investors are reluctant to fund last mile efforts, that is, broadly guess that in average overall that market, really “competitive” or not, is so well supplied that profits stand to be low.So, sure, as a last mile owner, to get more earnings, might help to have, overall, fewer last mile providers. So, e.g., are seeing mergers.In my neighborhood, everyone has connections via all three of twisted pair, coax, and wireless. So, maybe among the three, no one of them will be making big margins.

    3. Guesty McGuesterson

      well that was awfully Randian

    4. Rob Larson

      Sounds like you are describing SpaceX’s ambitious plans to become an internet service provider via a network of 4000 satellites (4x the current total of ALL active satellites). Which will be utterly transformative in rural / low-populated areas, but much less so in high population areas due to the bandwidth constraints that arise from many people accessing the same satellite. Fortunately it’s the rural areas that need broadband options the most.I like your ending prediction “And then another will rise”

    5. Michael Elling

      @Andy your analysis abjures the role government played over the past 30-50 years to get us to this point. What about equal access? What about number portability? What about mandated dual-mode devices? What about Part 15 spectrum (aka Wifi)? What about Computers 2 & 3? The list goes on and on. But you may be coming to your conclusion from the fact that more often that not the law of unintended consequences often took over in each instance. And that’s because governments are not planners as they do not know how to take risk and allocate capital efficiently.That said, the role of government is extremely important in all sorts of networks (transportation, utilities, etc…). And it is extremely important in networks. What we haven’t completely understood in the past 102 years is that we could have forced layer 1 and 2 resale from the start. But we took a wrong fork in the road in wireline markets in 1913 and again in wireless in 1934.The role of the government should be simply this: 1) mandate interconnection (even all the way out to the edge; aka Carterphone); and 2) monitor, guide and enforce settlement fees. Not set them. What the “IP stack” and net neutrality crowd are missing is that balanced settlements north-south (between apps and infrastructure) and east-west (between agents; be they end to end or core to end, etc…) serve as price signals and incentives. The “IP Stack” doesn’t have inherent price signals. That’s why it can’t migrate from IPv4 to IPv6 in 2 decades. Just one of many things the IP stack can’t do well (including security, etc…)But price signals and incentives are necessary for new (end to end) service creation that’s low cost and ubiquitous. These balanced settlements will reflect not only the marginal cost at the boundaries (horizontal and vertical), but also value transference. What few realize is that the bill and keep (settlement free) model actually fosters monopolies (which is what we have now at the core and edge) and impedes new entrants.I’ve said this to Brad. He’s not understanding the math of (inter) networks and taking the correct lessons learned from universal service and intercarrier compensation and applying them to a competitive informational stack. Right now there is no concerted way that the core can drive broadband adoption at the edge. And selling our privacy is not the answer.

      1. JLM

        .Extra points for “abjure.”JLMwww.themusingsofthebigredca…

        1. Michael Elling

          It’s weird, but sometimes words come from the dark recesses and I have to look them up to be absolutely sure I got it right. This was one of those instances.

          1. awaldstein

            I actually used concupiscent today–the ghost of Wallace Stevens touching this ex English major who can still recite his poems.

          2. Michael Elling

            Good you know the source. I’ve been bingeing on Blacklist lately and wondering if it isn’t a word used by Raymond Reddington.

          3. Matt A. Myers

            Had to look it up — good word.

  8. William Mougayar

    You can replace Canada there and the spirit of the problem and situation are fairly similar. We also need smarter politicians that believe in light touch regulation, and are not easily influenced by the incumbents lobby.

    1. Michael Elling

      Oh Canada! The land that introduced 10 cent digital wireless pricing in 1996! Where did it all go wrong?

      1. William Mougayar

        Exactly. We were leaders up until then.

  9. Ian Smith

    In the UK we have a lot of competition and low-cost options. The twisted-pair phone monopoly (BT) got opened up and many ISPs rent their line equipment. One ISP, TalkTalk, even has their own line cards in the local exchanges and offers the cheapest service. But we still have the problem of getting decent speeds to people. That is still owned by a monopoly (BT) and ageing copper cables.In the 1980s BT was ramping up to economically install fiber everywhere and got stopped by the politicians on thin grounds of it being a monopoly. Just think where we could have been now…In the 1990s the coax cable TV providers dug up a lot of streets and installed a digital coax service in cities where it was economical to do so. This offers much better speeds.I live in a village with 1.5MBPS twisted-pair broadband. The best speed I can get is with 4G wireless. That’s probably the only way forwards unless the government upgrades the infrastructure. It feels a bit like the railways.

    1. Michael Elling

      To get investment in the lower layers we need settlements that run north-south between layers and to get investment out to the edge we need settlements that run east-west (WAN/MAN/LAN/PAN) between agents/actors. The IP Stack lacks necessary price signals and incentives. The value is at the core and top in networks. These issues were not raised during the net neutrality debate because even the old telco model is going bill and keep (aka settlement free). Silo’s everywhere! Every agent is an island. Wonderful!

  10. Jim Kessler

    Should Facebook be lightly regulated so other companies have access to its members, data, pipeline, etc? Why do we assume that competition is a good thing?Dramatically better alternatives are much more likely to appear when there is a lack of competition and the margins are fat. When there is lots of competition, you are more likely to see a lack of great new alternatives (not more).

    1. Jamie Friedrech

      >Dramatically better alternatives are much more likely to appear when there is a lack of competition and the margins are fat. When there is lots of competition, you are more likely to see a lack of great new alternatives (not more).Evidence?

      1. Jim Kessler

        Land line phone companies with fat margins enabled the opportunity for others to make massive investments to create a wireless phone system.The last mile cash cow is creating the same opportunity and as we all know there are more than a few companies that are working to eliminate the need for the last mile to exist.Where is the evidence that relentless competition within an industry produces dramatic or even significant improvements especially ones that lead to dramatic change?If the margins are thin, there isn’t enough capital to pay for even mediocre improvements.

        1. Jess Bachman

          I think this whole startups thing is a good example of crazy innovation due to low barriers.Sure, if you dam up a river, it will overflow eventually… but who wants to wait 50 years for the next innovation to flood in?

        2. Peter Radizeski

          You cellular players are mostly telcos. The only notable exception is T-Mobile which was a roll-up. US Cellular, Sprint, VZ,and AT&T are ILECs. Only a couple of new entrants acquired by the Big 4. The ILECs used profits from landline – which had NO competition despite UNE-P CLECs until cable came along in 2000 – to fund the cellular networks, which they basically have a monopoly on.

    2. SubstrateUndertow

      Really !Your gonna run with that as your historical summary ?

      1. Jim Kessler

        What is your counter argument?

        1. SubstrateUndertow

          The historical record !

    3. pointsnfigures

      Yes, you have a choice whether to use Facebook or not. You have a choice whether to use Google or not. You often don’t have a choice over your internet provider (or your health care provider)

  11. Salt Shaker

    The telco and cable industry stranglehold on Washington is as strong as Crazy Glue.No incentive to foster competition when legal monopolies run amok.With respect to regulation–light or heavy–I think a quote from the late great Yogi Berra (RIP) sums it up nicely when he said “we made too many wrong mistakes.”

  12. Maya Wiley

    Competition yes! And regulation that protects it yes! If we protect consumers and incentivize equity, I’d call that appropriate regulation.

  13. Jess Bachman

    Can’t we just put broadband in the blockchain already.. jeez…

  14. pointsnfigures

    I think net neut will limit, not expand competition. Already, we are seeing the big telcos invest less in their networks. (http://www.wsj.com/articles… I 100% agree with you that competition is the way to expand service and lower prices. It extends not just to broadband, but to health care and everything else. Demand curves slope down.Ironically, in this day and age where network beats hierarchy, and everyone is a node on a network; networks not the government are better regulators and can do it far more efficiently (and cheaply). Good functioning networks are essentially a market and markets allocate resources better than any bureaucracy.

    1. Peter Radizeski

      Light regs, heavy regs; the Duopoly does not compete. VZ did a co-marketing deal with the Cablecos in exchange for their spectrum (SpectrumCo deal). The cablecos don’t over build over each other. Cablecos are awful wholesale partners to CLECs. ILECs don’t compete across regions in broadband. WISPs, CLECs and PCOs are the only third parties building out fiber to the home/MDU/MTU.Investment is down for a couple of reasons: (1) costs for FTTx have decreased; (2) CAF and other funds are being used in place of ILEC CAPEX. (3) telcos haven’t deployed a plan yet to take back broadband market share from cablecos. Most ILECs have given up competing against DOCSIS 3.0. Until the vendors prove g.fast or some DOCSIS killer, investment will remain low.

    2. Michael Elling

      In networks such as communications and transportation there are natural layer 1 monopolies brought on by rights of way issues. Wireless gets around a lot of these, but we auctioned off spectrum (raising everyone’s monthly bills by 20-30%) and didn’t mandate interconnection. Why does wifi-offload account for 70% of the broadband consumption on smartphones? Government does have a very simple role: mandating interconnection and settlements between the nodes (or networks). What’s missed by the net neutrality wonks is that settlements are critical because unlike transportation and other utility networks, communication networks are real-time and two-way. The others are all store and forward. So simultaneous end-to-end change and investment is necessary to make the process democratic. This only comes through price signals and incentives; aka settlements.

    3. PhilipSugar

      I don’t know. Do you remember the Ma Bell days?

      1. pointsnfigures

        I do. One regulated monopoly. Then they broke them up. Remember how people complained? They lobbied, consolidated, and now it’s one big more regulated oligopoly. We’d be better off letting it become the wild west with very little regulation. Would be interesting to see where it went. Coase Theorem on steroids. It would initially be messy, but at the end it would be better and work.

  15. DJL

    If memory serves, this is because the big integrated players where the only ones willing to make the massive investment required BEFORE there was a huge broadband market. (Remember AOL and dial-up?) If there were smaller players, they got “integrated” the same way that any company buys up channel players. So trying to create a certain market situation is hopeless when you have free markets.Instead, try simple items like rejecting the AT&T purchase of DirectTV. Now my great customer service will go in the tank. Focus on the customer – and the market will take care of itself.

  16. iggyfanlo

    My experience tells me that more and more of the world falls into the public good, common good or club goods category. As a derivatives trader, I never felt that way, but the various financial crises and the one way option that financial firms have taught me that this was truly a public good… also true of fixed wires…Yes, if/when we get very high speeds from wireless then that categorization goes away, but until that time, regulation, size limitation, legally required and enforced sharing (with all its political flaws) needs to move forward

  17. JLM

    .Innovation does not have a finish line.Left alone, entrepreneurs will continue to innovate — FOREVER. That is exactly what is supposed to happen. Continuous improvement, innovation — says the guy with 5 different providers of 1Gig Internet service in Austin by God Texas wherein once upon a time there was none.The problem with regulation is that it creates a finish line wherein the big boys rally at the winner’s circle and beat their chests and spend all their time trying to keep everyone else out.The big boys love regulations because they are able to impact their drafting — they actually draft them as in the case of Obamacare which was drafted completely by outsiders — and then their application until the winner’s circle becomes a moated castle and nobody else can ever get in.”Smart light regulation” is like a mild case of Ebola or AIDS.Innovation flourishes when there is no finish line.JLMwww.themusingsofthebigredca…

    1. Anne Libby

      Huh, and I thought that Ebola, AIDS or cancer happens when a disease grows without constraint.

      1. JLM

        .And thereby unmasking the intent of my analogy. Thank you.There is no such thing as light smart regulation — it just keeps growing because the next guy thinks HIS little packet of rules are also light, smart, and necessary.Madame Obvious, you.JLMwww.themusingsofthebigredca…

        1. Anne Libby

          lol(I recently read that only Olds used “lol” any more. lol.)

    2. Twain Twain

      EXACTLY THIS: innovation flourishes when there is no finish line.To be sure, great inventors understand there are milestones and they have to SHIP product.However, during the innovation process, there’s constantly new tools and information that emerge which enable inventors to make products even better.So there’s never a finish line because inventors don’t think in terms of limits and bounds like legislators.They think in terms of quantum leaps and not in terms of lines, anyway, haha.

    3. Matt A. Myers

      Until the end of time.

  18. T McGinn

    I think this would look a lot like the current state of deregulation in the retail electricity industry. I would argue that deregulation there has been mostly beneficial to consumers but that the lack of product innovation has been mostly disappointing.Does opening up the plumbing network in the broadband space allow for competition on the basis of quality of service? I don’t know enough about broadband to compare the infrastructure to the electricity grid. In electricity, the service quality is commoditized and retailers compete for customers mostly with gimmicky contract terms, loyalty rewards programs, multilevel marketing channels, affinity networks, and other branding efforts.Browse an industry news source and you will easily find multiple stories of customer manipulation, actions takes by consumer protection bodies, employment disputes involving contract sales people, allegations of fraud, etc…http://energychoicematters….Bloomberg has a story today about a handful of Texas retailers searching for value-add components to battle persistently high churn rates:http://t.co/5ehgy3QDWbMy first reaction to the excerpt from the WH release was to think that the underserved communities are the least likely to benefit from deregulation. Navigating the competitive landscape in the Texas electricity market without significant industry knowledge is essentially impossible. My fear would be that the broadband industry would end up the same space.Here is what Texas consumers face (307 offers in the Houston area):http://www.powertochoose.org

    1. Michael Elling

      Electricity markets have mostly been a 1-way affair (except for very large users who could build their own power plants and sell into the grid). That’s all changing now and furthermore, it is going mobile with storage. So electricity and communications are looking a lot more similar than not. Neither industry is well positioned for the changes coming. It starts with managements understanding who the customer is and then what position to occupy; dominate a horizontal layer or niche or develop vertically complete solutions.


    You lost me at regulation. Forget it.

  20. avifreedman

    Couldn’t agree more.In 1996 we had a peering panel at ISPCON. John Curran was there. At the time he was CTO of BBN, the hardest ISP to get peering with at the time.After people vented, he stood up and said:”I know you think I’m evil and someone should do something to make peering easier or more open. I just want you to consider that it’s now 1996, 4 years after the spread of the commercial Internet and 2 years after the web browser kicked things into gear. We now have 2 ISPs in every city of this country selling unlimited access for < $20/mo.What do you think we’d have if the government was specifying or running things?”In South Korea it went down differently but I agree, it’s the private market most likely to make a dent in the US.We just need a national broadband competitor with Uber-level rough and tumble. It’s so much harder to fight municipality and state government stalling and lobbying, and much more capital intensive than the modem business.Would love to see USV fund folks working in this space, good luck…

  21. TeddyBeingTeddy

    chris Dixon would love the telecom full stack!

  22. Michael Elling

    It is clear that the vertically integrated business model is doomed. But to make the transition smooth incentives are necessary to drive broadband adoption from the core that takes into account the value at the core and the rapid obsolescence everywhere in the stack. Settlements are necessary and more often than not the larger player needs to pay the smaller player a disproportionately larger relative amount to capture a larger absolute amount of network effect. The business model math argues against settlement free (bill and keep) interconnection for both the core (AGFA) and edge providers (MSOs and MNOs).

  23. Sam

    Codified property ownershipEnforcement of contractsLaws against use of force, fraud, theft of propertyRules that permit exchange of propertyThese are some of the necessary conditions for a free market to operate effectively. And yes, they are regulation. Some regulation is necessary for efficient functioning market economy.Beyond that, regulation can be important for minimizing negative externalities (like environmental pollution) and protecting and promoting public goods (like, arguably, access to broadband Internet).It doesn’t strike me as very helpful to knee-jerk react with “regulation bad” when there is something important at stake here.Unintended consequences scare the crap out of me when we’re considering new regulation. “The road to hell is paved with good intentions” and all that. But if you believe it’s a public good, then you have to try. The head-in-sand alternative gets us nowhere.

    1. Stephen Voris

      Knee-jerk “regulation bad!” is about on par with knee-jerk “there ought to be a law!” Neither is doing more than preaching to the choir.One wonders if regulations might benefit from a separation of their intents and their specific prescriptions (regulating the regulations, as it were); one of the major dangers with regulation is that it doesn’t get fired for nonperformance… rather, it’s quietly ignored but left on the books.

  24. virme

    I wonder how long before we start discussing similar models for other players in tech industry. We only have two big players ($FB and $GOOG) in user and data ownership. At some point they will they not be interested in innovation

  25. Stuart Willson

    CLECs failed for a lot of reasons. UNE-P is one, but another – for those who invested in fiber – was being early. And another – for everybody else – was not being able to leverage a very profitable business (video) to invest in the potential fiber use case (today: 100%+ of NPV). But today, with bandwidth seemingly in short supply and demand with a clear path to greater need, I’m curious what sorts of *light* regulation you think will incentivize the owners of spectrum/fiber to open networks to competition. I was looking through the comments and don’t see any suggestions. There must be something more creative than the obvious three options: 1/ force incumbents (presumably cable and ilecs) to lease their capacity to entrants, 2/ incentivize customers to use competing providers (making the overbuild economics more attractive), or 3/ incentivizing overbuilding (in the air or in the ground).

  26. Clyde F. Smith

    Another important piece of the puzzle is stopping state level government from stopping municipal level development of broadband services. That’s been a big problem from Republican-dominated state legislators in both NC and Tennessee for just a couple of obvious examples. Things are improving in TN but the movement forward has been almost totally blocked in NC for municipal broadband. The pockets of big business have already blocked a lot of communities from getting the broadband they want under local control.

  27. Savethewater

    Your right on the money! Competition and Choices is what consumers need.

  28. Clyde F. Smith

    Chattanooga boosts citywide broadband capacity to 10 gigabitshttp://www.timesfreepress.c…