Net neutrality: what’s a libertarian to do?

One of my commenters asked, rather plaintively:

You mentioned net neutrality. I’ve read about this, and the opposition to it. I’ve read about this, and the opposition to it. As far as I can tell, net neutrality is more supported by liberals/democrats, while the opposition is made up more of conservatives/republicans. But for the life of me I can’t figure out which is the the more libertarian position.

Your confusion is entirely reasonable. I’ve hung out with network-neutrality activists and tried to give them what I thought was useful advice. Their political fixations didn’t permit them to hear me. Here’s a summary of the issues and one libertarian’s take on them.

Here’s where it starts: the wire-line telcos want to use their control of the copper and fiber that runs to your house to double-dip, not only charging consumers for bandwidth but also hitting up large content providers (Google, Amazon, etc.) for quality-of-service fees. There’s another question that gets folded into the debate, too: under what circumstances the telcos can legitimately traffic-shape, e.g. by blocking or slowing the protocols used for p2p filesharing.

It is not clear that the regulatory regime under which the telcos operate allows them to do either thing. They haven’t tried to implement double-dipping yet, and they’re traffic-shaping by stealth and lying about it when they get caught. What they want is a political green light to do both.

Let it be clear from the outset that the telcos are putting their case for being allowed to do these things with breathtaking hypocrisy. They honk about how awful it is that regulation keeps them from setting their own terms, blithely ignoring the fact that their last-mile monopoly is entirely a creature of regulation. In effect, Theodore Vail and the old Bell System bribed the Feds to steal the last mile out from under the public’s nose between 1878 and 1920; the wireline telcos have been squatting on that unnatural monopoly ever since as if they actually had some legitimate property right to it.

But the telcos’ crimes aren’t merely historical. They have repeatedly bargained for the right to exclude competitors from their networks on the grounds that if the regulators would let them do that, they’d be able to generate enough capital to deploy broadband everywhere. That promise has been repeatedly, egregiously broken. Instead, they’ve creamed off that monopoly rent as profit or used it to cross-subsidize competition in businesses with higher rates of return. (Oh, and of course, to bribe legislators and buy regulators.)

Mistake #1 for libertarians to avoid is falling for the telcos’ “we’re pro-free market” bullshit. They’re anything but; what they really want is a politically sheltered monopoly in which they have captured the regulators and created business conditions that fetter everyone but them.

OK, so if the telcos are such villainous scum, the pro-network-neutrality activists must be the heroes of this story, right?

Unfortunately, no.

Your typical network-neutrality activist is a good-government left-liberal who is instinctively hostile to market-based approaches. These people think, rather, that if they can somehow come up with the right regulatory formula, they can jawbone the government into making the telcos play nice. They’re ideologically incapable of questioning the assumption that bandwidth is a scarce “public good” that has to be regulated. They don’t get it that complicated regulations favor the incumbent who can afford to darken the sky with lawyers, and they really don’t get it about outright regulatory capture, a game at which the telcos are past masters.

I’ve spent endless hours trying to point out to these people that their assumptions are fundamentally wrong, and that the only way to break the telco monopoly is to break the scarcity assumptions it’s based on. That the telecoms regulatorium, far from being what holds the telcos in check, is actually their instrument of control. And that the only battle that actually matters is the one to carve out enough unlicensed spectrum so we can use technologies like ad-hoc networking with UWB to end-run the whole mess until it collapses under its own weight.

They don’t get it. They refuse to get it. I’ve been on a mailing list for something called the “Open Infrastructure Alliance” that consisted of three network engineers and a couple dozen “organizers”; the engineers (even the non-libertarian engineers) all patiently trying to explain why the political attack is a non-starter, and the organizers endlessly rehashing political strategies anyway. Because, well, that’s all they know how to do.

In short, the “network neutrality” crowd is mainly composed of well-meaning fools blinded by their own statism, and consequently serving mainly as useful idiots for the telcos’ program of ever-more labyrinthine and manipulable regulation. If I were a telco executive, I’d be on my knees every night thanking my god(s) for this “opposition”. Mistake #2 for any libertarian to avoid is backing these clowns.

So, what are libertarians to do?

We can start by remembering a simple truth: The only substantive threat to the telco monopoly is bandwidth that has been removed from the reach of both the telcos and their political catspaws in the regulatorium. Keep your eye on that ball; the telcos know it’s the important one and will try to distract you from it, while the “network neutrality” crowd doesn’t know it and wastes most of its energy self-defeatingly wrestling with the telcos over how to re-slice the existing pie.

Go active whenever there’s a political debate about “unlicensed spectrum”. More of it is good. Oppose any efforts to make UWB (or any other technology that doesn’t cause destructive interference) require a license anywhere on the spectrum. If you are capable, contribute to the development of mesh networking, especially wireless mesh networking.

Oh, and buy an Android phone. As I noted in my immediately previous post, Google is our ally in this.

UPDATE: I’ve summarized the history of the Bell System’s theft of the last mile here.

80 comments

  1. > ad-hoc networking with UWB to end-run the whole mess until it collapses under its own weight.

    sorry, won’t work, and the telcos aren’t worried about it. (AT&T spent $2B on “Project Angel” just to tip the market.)

    What *will* work is community-owned, bond-financed FTTx.

  2. As Vadim Antonov has repeatedly said, the key to insufficient bandwidth is not better QoS (QoS fees or traffic-shaping). No, the key to insufficient bandwidth is …. more bandwidth.

  3. Yeah, as much as Jim and I can be at loggerheads, the key to the last mile problem is municipal or cooperative-owned FTTH (I say “go all the way to the house”).

  4. > Go all the way to the house.

    certainly, though GigEnet is likely enough to any individual residence.

    but talking about bandwidth without ensuring connectivity is cart-before-horse.

  5. “Oh, and buy an Android phone. As I noted in my immediately previous post, Google is our ally in this.”

    Not until Google can be proven to:

    1. Actually release a secure product that doesn’t do incredibly insane things, such as run textpad entries as root shell commands.

    2. That they won’t become the evil we seek to replace them with. I keep getting the feeling that Google is the equivalent in all of this of the USSR in World War II, an ally of necessity, not a true friend. If it makes them a profit, they’ll eventually turn to things that will probably have many screaming in the wake of them. It’s what corporations do.

  6. I think sooner or later the Internet has to grow out of the whole “something piggybacking upon the phone lines” paradigm…

  7. My reason for supporting network neutrality, in whatever form it can be accomplished, is simpler. I believe in the free market. Unfortunately, in the case of the backbone providers that want to soak the Googles and Microsofts of the world, there’s no lever for the market to push on. I can’t tell my traffic “I don’t like AT&T’s charges for transit traffic that want good service, so don’t route over AT&T”. In the absence of such a lever, other answers – be they regulatory or technological – are necessary. I’m not convinced that unlicensed spectrum is the answer, simply because there will always be required infrastructure that someone must pay for. The details aren’t as important as the underlying concept: the Internet must be a level playing field for everyone, from the individual looking to make a splash all the way up to Google.

  8. > but talking about bandwidth without ensuring connectivity is cart-before-horse.
    Mmmmm, if I can run 2.5 miles of fiber down my road to the state highway, we’re then within shooting range of the CO, cable goes down the state highway, there are three company’s fiber runs right there, AND the local ISP is (slowly) running fiber up the highway. Plenty of options once you have a home-run patch panel with a couple of dozen houses.

  9. Jay, then you’re talking about a different NN than the “activists” are. They want to force the last-mile providers to do certain kinds of carriage and/or prohibit other kinds of carriage. I don’t want to force them — I want to GO AROUND them.

  10. The folks who treat NN as a political problem drastically underestimate the consumer. An “internet” connection where google, Amazon, MySpace, Facebook, a third-party email application (corporate emaIl servers count), or a VPN feel slow is not going to be perceived by the consumer as a valid fulfillment of their contract with the ISP, whether it technically is one or not, and they’ll stop paying for it and go to whoever will supply it to them.

    What’s kept Microsoft’s almost-monopoly going so long isn’t so much market power as Microsoft’s own understanding that to stay on top, they have to continue supplying their users with at least a minimally acceptable experience. It can be (and has been) garbage from an engineering perspective, but as long as it doesn’t induce enough pain to cause users to stop buying upgrades, most people won’t leave. At least up until Vista, Microsoft’s products met this “minimal acceptability to the average OS buyer” criterion.

    But a non-NN-compliant broadband connection which penalizes either a publically-popular service or a critical-path corporate infrastructure application for refusing to pay protection money does NOT meet this criterion, and won’t ever again. It violates the commonly-accepted definition of what an “internet connection” _is_, and as long as there’s any competition at all, the market is going to punish any moves in that direction very harshly.

    Expanding the power of politicians and bureaucrats is _never_ the right answer to a problem which has ever been solved in any other way. The market has been solving the NN problem since years before the mainstream population had ever heard the word “internet”. There is no compelling evidence that it cannot continue to do so, and plenty of compelling evidence that it can. All contrary statements are mere conjecture, which can’t be allowed to trump the facts on the ground.

  11. “I have been involved with Indymedia since the autumn of 2000” – well, that sort of automatically gives me +5 on the saving throw against believing this stuff…

  12. Shenpen:

    I nearly sprayed coffee on the terminal after reading that. Shenpen is attacked by meme spell. (rolls dice) makes saving throw!

  13. The question in my mind is how is FTTH going to be financed? Somebody has to pay for it.

    It is all the Last Mile question. We want fiber, and the telcos claim that the only way they can profitably roll it out is to double-dip as Eric calls out.

    Some new subdivisions are already FTTH, in which case the monopoly problem is solved. What to do about existing homes that want fiber but don’t want the telco to have a property right to it?

  14. The recent FCC ‘white spaces’ decision will have an incredible effect
    on the availability of internet service that bypasses the cable and
    telephone monopolies, at least on the scale of wi-fi, and most likely
    greater because the white spaces are lower in the RF spectrum. The
    real way to beat the monopolists is with unregulated spectrum and
    that’s why Google and Microsoft and the other biggies pushed so
    hard to open up the UHF TV band to non-TV operations. When the
    National Association of Broadcasters recruited Dolly Parton to oppose
    the action, it was a hilarious act of desperation.

  15. I always thought that technological measures, such as mesh networking and using “white spaces” for Net access were much better than regulations anyway, since technology moves much faster than legislation. Unfortunately, most “free market” advocates side with the telcos in opposing Net neutrality. It seems that capitalists are the useful idiots this time around.

  16. What is the difference between the US and Europe/Asia where they seem to be years ahead of us with regard to bandwidth?

  17. @Jay: What is the difference between the US and Europe/Asia where they seem to be years ahead of us with regard to bandwidth?

    Most of the difference is that Europe and the Pacific Rim have lower dispersion of population over smaller areas than the U.S., so their deployment problem is substantially easier. Some of it is the U.S. telcos. As I wrote in the post, they’ve been successfully gaming the political system to evade having to do the universal broadband deployment they were supposed to in exchange for continued government sponsorship of their oligopoly.

    @DD: It seems that capitalists are the useful idiots this time around.

    No, the conservative corporatists. The “capitalists” are libertarians like myself and Russ.

  18. While I agree that you and Russ Nelson support capitalism (you, in-particular, anarcho-captialism while Russ seems to be in a free-market frame of mind), being a “capitalist” means that you *own* actual *capital*.

  19. Although I agree with your analysis of the teleco’s position, I wonder if your solution comports with the basic underlying principles of libertarianism, nay, capitalism. Unlicensed spectrum is essentially a commons, and we are no doubt familiar what happens to the commons. What is needed is property rights in the spectrum. There are some very messy approximations to this, but the government certainly has not allowed unfettered use of the spectrum that is bought. It is more like a highly restricted leasehold (like having logging rights in a national park.)

    Property rights are almost always the solution to scarcity, because property owners have an incentive to improve their property, and use it as efficiently as possible. The solution is to split the spectrum up by frequency band, and geographical area, and sell the whole thing off. Perhaps keep some for government use, but put it in the hands of speculators who will make the maximum use of the bandwidth.

    No doubt unlicensed spectrum has been a big success for WiFi, garage door openers and cordless phones. But the success there is largely due to the fact that the short range means that it is a defacto property right: I control what WiFi signals get into my house since it is my house and so I can control what transmitters are allowed to be plugged in. There is some leakage, but the power caps (and non governmental standard protocols) prevent that from being a problem. Same goes in Starbucks and Borders.

    However, if you allow it out of direct proprietary control, you get all the mess you expect with the commons (consider municipal WiFi for example.)

    Although I agree with many of your points, I don’t believe that more unlicensed spectrum is the optimum answer. (That is not to say it is not the best that can be expected when factoring in the realities of politics.)

  20. >@DD: It seems that capitalists are the useful idiots this time around.

    >No, the conservative corporatists. The “capitalists” are libertarians like myself and Russ.

    I was referring to opinions offered by libertarians/Objectivists that argue a property rights case against Net neutrality while ignoring that the baby bells are a product of government regulation.

    The big reason that regulation won’t work on the Net is its promiscuously international nature. Any attempt by one nation would be ineffective in the rest of the world. Remember how we don’t see any content that is offensive to the Chinese government? ;-)

  21. >Although I agree with your analysis of the teleco’s position, I wonder if your solution comports with the basic underlying principles of libertarianism, nay, capitalism. Unlicensed spectrum is essentially a commons, and we are no doubt familiar what happens to the commons. What is needed is property rights in the spectrum.

    Until relatively recently I would have agreed with you. I used to rant at my friends about how the 1934 power grab by the Feds aborted a property-rights system for radio frequencies in given broadcast areas, treating them analogously to homesteaded land with interference being interpreted as trespass. Yes, the courts had been heading in this direction.

    More recently, I have been convinced that frequency scarcity is an artifact of stupid radios. That is, if you have software-defined radios doing spread-spectrum packet protocols, it’s possible for two radios to negotiate a channel-hop sequence and get data through even under conditions of extremely high and time-varying noise from other transmissions. As a side effect, communication is unjammable. Yes, there’s an aggregate upper limit of total over-air bandwidth, but it’s ridiculously high and not a practical constraint.

    Under these circumstances, spectrum is effectively non-rivalrous, and you therefore don’t need any analogue of property rights. The radios will stay out of each others’ way in order to minimize power expenditure per transmitted byte.

    Final note: This is not a blue-sky possibility. It is nearly certain that this is the way U.S. military radios already work. What’s novel is that today the hardware to do it can be manufactured in bulk at near-throwaway prices.

  22. >I was referring to opinions offered by libertarians/Objectivists that argue a property rights case against Net neutrality while ignoring that the baby bells are a product of government regulation.

    I’m probably not going to surprise you by telling you that I think these people are either dupes or frauds. I’m pretty sure Russ will agree with me on this.

  23. There’s another leg of NN advocacy that may be influencing the ‘organizers’ to which ESR alludes. In certain quarters, NN is regarded as primarily a form of content ‘neutrality’, meaning that network providers should not be able to ‘discriminate’ against ‘socially conscious’ media and services delivered over the internet by charging them market rates. The logic here is that since network providers are for-profit corporations and most of their clients are for-profit corporations, they should be obliged to give capacity to not-for-profit entities as a public service (i.e. commercial rates impede access by non-commercial entities). So, for instance, network providers should subsidize services like Current TV.

    You’ll encounter this argument frequently in the non-profit film, television, and arts communities. My company interacts w/ these folks on a frequent basis, and so I’ve had an opportunity to track their understanding of the NN issue since its inception. When NN first became a prominent issue, this community immediately gravitated to an analysis of the conflict that is based on the assumption that the alternative to NN is active and deliberate censorship – that network providers didn’t intend to be ‘content neutral’. To be fair, some of the early advocacy by Google and other companies promoting NN did employ ‘neutral’ in this sense – so there was some justification for their believing that this was the case. At this point, their support for NN is predicated on the expectation that it can be leveraged to provide subsidies to their organizations, to create a sort of PBS for the internet.

    Here’s an example of what I’m going on about.
    http://foureyedmonsters.com/save-the-internet/ – this is not a parody.

  24. >>I was referring to opinions offered by libertarians/Objectivists that argue a property rights case against Net neutrality while ignoring that the baby bells are a product of government regulation.
    >I’m probably not going to surprise you by telling you that I think these people are either dupes or frauds.

    Like a lot of people with libertarian leanings, I did go through an Ayn Rand phase. I like her defense of free markets, though I’m not too sure about that “osycho-epistemololy” business. She was a brillian fiction writer, even if her characters came dangerously close to being Mary Sues.

  25. >I like her defense of free markets, though I’m not too sure about that “osycho-epistemololy” business.

    I presume you mean “psycho-epistemology”, and if so your instincts are good. Rand was a brilliant moral critic but an embarassingly, cringe-makingly bad epistemologist. I might blog a detailed takedown of her epistemology sometime; it’s not a difficult dissection for anyone who knows even a smidgen about 20th-century analytical philosophy and the Wittgensteinian or Korzybskian analysis of language. Topic sentence: she mistakes Aristotelian logic for a feature of reality rather than a contingent artifact of language, and from there much nonsense flows.

    To any Randites listening: Calm down, OK? Ayn was dead right about the whole altruism thing. That’s the essence; the bogus epistemology isn’t necessary to it.

  26. >I presume you mean “psycho-epistemology”, and if so your instincts are good. Rand was a brilliant moral critic but an embarassingly, cringe-makingly bad epistemologist.

    Thanks for calling attention to my fat-fingering, Eric. ;-)

    Rand’s neologisms such as “psycho-epistomology” do sound impressive to 19-year-olds who have read nothing about philosophy. Actually, my route out of Randian epistemology was Sausurrian semiology, which doesn’t seem too far off from Wittgenstein or Korzybski.

  27. @Jessica: are colors scarce? If colors are not scarce, and colors are a wavelength, and radio spectrum is comprised of those same wavelengths, then how can spectrum be said to be scarce?
    @Mark: the cost of FTTH is approximately the same as a new high-def TV. People can afford to own their own TVs; why not own their own fiber?
    @ESR: I told Kragen Sitaker that we should give the telcos a deadline for provisioning FTTH, or give free space on the poles to somebody who will. He said “I’m shocked you would say that. It shows me that my Russ Nelson model is grossly inadequate!”
    @Matt: the difficulty is that the people pushing non-neutrality are sitting between the ‘net and people’s houses. Not many providers there; not much competition. Some; not enough.

  28. From what I read so far, the only Android powered phone available is currently from T-Mobile (in some parts of Europe “affectionately” known as “Hitler-Telecom”, for their extremely abusive monopolistic practices, and treating their customers like the dirtiest filth on planet Earth).

    What I also understood is that the only way to get a working Android phone is to buy one with a calling plan from the said telecom.

    So if I understood correctly, I absolutely refuse to buy Android powered anything from T-Mobile until I am free to choose any SIM card, from any telco provider (unlocked phone), and especially, I refuse to buy an Andorid phone from T-Mobile until I can put in a prepaid SIM card from any provider, and it will just work.

    I won’t stand for a locked-anything in a mobile phone; and until such time as completely unlocked, Anrdoid powered phones flood the market, I will boycott any such SIM locked device.

    Nobody orders me how I am to use a device I pay for. I gave uncle Steve (of Apple Computer) the finger, and I’m most certainly not going to put up with any antics from “Hitler-Telecom”.

  29. > In effect, Theodor Vail and the old Bell System bribed the Feds to steal the last mile out from under the public’s nose between 1878 and 1920

    Eric, could you please elaborate on this? On the one hand I (and possibly others) lack information on this item of American history. On the other, it might help clarify your point on telcos for those less likely to accept libertarianism at face value.

  30. @Russ:
    > are colors scarce? If colors are not scarce, and colors are a wavelength, and radio spectrum
    > is comprised of those same wavelengths, then how can spectrum be said to be scarce?

    Spectrum ‘scarcity’ is driven by the current models and definition of ‘interference’. The thing is, ‘interference’ is an artifact of receiver design. Not unlike the duality of light (is it a ‘particle’ or a ‘wave’? Its *both* until you observe it), there is *no* interference until some receiver attempts to recover a signal.

    But of course, we can do far, far better than current receiver (PHY) design (and make a better MAC-layer protocol too.)

    I refer the readers of this blog (again) to TIm Shepard’s MIT thesis. http://publications.csail.mit.edu/lcs/pubs/pdf/MIT-LCS-TR-670.pdf

    Note that its all about “bandwidth” arising out of *dense connectivity*.

    But otherwise, I’m singing in harmony with Russ’ comments here. Free poles would be a step in the right direction.

    See also: http://www.frankston.com/public/?name=FSM

  31. ” “Hitler-Telecom”, for their extremely abusive monopolistic practices, and treating their customers like the dirtiest filth on planet Eart”

    Partly because of that and partly because their old name was Deutche Telekom and for some people these jokes never get too old :-/

  32. >Sausurrian semiology, which doesn’t seem too far off from Wittgenstein or Korzybski.

    I agree — but somewhat grudgingly, because Sausurre was a sloppy thinker. Korzybski or C.S. Peirce would have been better value for your time.

  33. >Eric, could you please elaborate on this?

    Beginning around 1899 (and especially after his return from retirement in 1907), Theodore Vail successfully lobbied the government to give AT&T regulated monopoly status. His hook was universal access, especially for rural areas: “It is not believed that this can be accomplished by separately controlled or distinct systems nor that there can be competition in the accepted sense of competition.” The government bought this argument, and the agreement was formalized by the Kingsbury Commitment of 1913.

    For a critical analysis. see UNNATURAL MONOPOLY: CRITICAL MOMENTS IN THE DEVELOPMENT OF THE BELL SYSTEM MONOPOLY. Perhaps the most important point: Vail’s claims that monopoly was (a) a natural consequence of network effects and (b) required for universal service were both falsified by facts. What actually happened was that local telephone companies had begun springing up to service areas where AT&T thought the investment to lay lines couldn’t be justified. Some were so successful that they began taking customers from AT&T in more urban areas; at the time of the Kingsbury Commitment, AT&T’s market share was actually falling.

    The libertarian-minded howled, but the fix was in. Regulatory capture followed swiftly, accelerated by a temporary nationalization of the telephone industry during WWI; the theoretically “independent” local carriers never recovered, being reduced to satrapies of Ma Bell. There would be no challenge to the monopoly, ever, from within the regulatorium — has never been, to this day. The 1956 Consent Decree, the Carterfone decision, and the AT&T divestiture all arose from antitrust action.

    This is actually one of the classic cases that demonstrate how backwards left-liberals have it about monopolies. They are never — ever — the natural result of unregulated markets, needing to be heroically busted by government. In fact, the exact reverse is true; “natural” monopolies have an extremely short half-life in a free market, and need to be propped up by regulation and coercive market-rigging in order to survive.

  34. T-Mobile was Deutsche Telekom? Really?

    Huh. I didn’t know that. I actually wonder how I managed not to know that.

  35. > Yeah, as much as Jim and I can be at loggerheads, the key to the last mile problem is municipal or cooperative-owned FTTH (I say “go all the way to the house”).

    The number of municipalities/co-ops that resist the temptation to filter will be dwarfed by the number who see filtering as one of the reasons to provide FTTH.

  36. > The number of municipalities/co-ops that resist the temptation to filter will be dwarfed by the number who see filtering as one of the reasons to provide FTTH.

    And you put your finger on the exact reason that I break out in hives when I hear the phrase “municipally owned”. Local-government control of that chokepoint is no better than big-corporate control. There are obvious ways in which it could be worse.

  37. > T-Mobile was Deutsche Telekom? Really?

    Really.

    > Huh. I didn’t know that. I actually wonder how I managed not to know that.

    Prior to 2002, they called themselves “VoiceStream” in the US.

  38. T-Mobile still is Deutsche Telekom. You can find references to their overseas parent in the fine print of their contracts.

    As for competition in the last mile being insufficient…well, I agree, with the proviso that “insufficient” is defined by reference to the ideal environment we ought to have, rather than by reference to what is necessary to ensure de facto NN policies without the requirement of legislative or regulatory intervention. The competitive environment of today is far from perfect or even desirable, but it is nevertheless minimally sufficient to achieve that particular objective.

    We should definitely argue for more competition. But the competition we already have is, insufficient thought it certainly is by other standards, nevertheless sufficient to ensure network neutrality in practice. The evidence for this proposition is obvious. To wit, de facto network neutrality exists today, despite the evident desire of incumbent players to kill it and no regulatory or legislative opposition to doing so. Ergo, new legislation or regulation _must_ be regarded as unnecessary for the achievement of this objective.

  39. Eric, your argument is that due to technological advances that bandwidth is not scarce. To be honest I think that it depends on what one means by “scarce”. I’m not an economist so I don’t know if there is a technical definition of scarcity (I am sure there is), but it seems to me there are two things that matter: if I am using it, can you use it too? (This would be my argment against using the word “property” in the context of intellectual property.) Secondly, is there a practical upper limit on the amount available for use? (This would be my argument against property rights in, for example, air.)

    In terms of the use of the radio spectrum, I would argue that the answers to these two questions are no, and yes, meaning that it is in the realm of property. In fact, the very argument you make would make me more inclined toward a property system. Things like UWB, spread spectrum, and smart radios in general, run into insane problems when the radio spectrum is managed by a creaking old bureacracy where, for example, you still need morse code to get a ham license. To do UWB you need to convince the FCC (or substitute your local regulatory agency) to allow you to do it, meaning you have to convince a committee that is largely in bed with the very people who don’t want you to compete, and loose out if you do. And it is not just malice, it is also incompetence (the two most prominent attributes of government being malice and incompetence — thankfully they often cancel each other out: if they didn’t we’d be really screwed.) That is to say, the slow creaking machine is very slow to adapt to changes in technology. (This by the way is part of the design of our government, and a good thing too.)

    If you have a property system, you by options with the specturm owners you want to step on, and convince them that they get free money allowing you to do it, because it doesn’t mess with what they are doing already. UWB is difficult because it transmits so broadly, but spread spectrum and smart radios can allow you to tightly control what frequencies you use, and allows you a better negotiating position (because you don’t need everyone to sell to you.)

    Consider this: in the context of privately managed bandwidth such as what cell phone companies have (and in the sense I described before) “junk” spectrum used by WiFi: the amount of innovation in use has been unbelievable. The growth and efficiency in use of these spectra have been completely out of parallel with the FCC controlled bandwidth. Cell phones, for example, have managed to squeeze multiple orders of magnitude of calls into the same badnwith over the twenty or so years they have been seriously around. (Compare also what you can do with your cell phone and what you can do with your desk phone.) WiFi is just about as fast as standard copper cable. Why is that? Because we don’t have the FCC dictating how and what can be done in these bandwidths. Of course, this statement isn’t quite true, the FCC does dictate to a certain extent, but they are more hands off, and look at the explosion of innovation. In fact, the very technologies you mention are largely there because of deregulation of parts of the spectrum.

    In regards to Russell’s question: is there a scarcity of colors? The answer is of course there is, in the application we are talking about. If I want a room lit with a soft sunlight glow, and you want the same room lit with a red glow, we are going to interfere, so in that sense there is a scarcity of colors or at least in the use of colors which is what really matters.)

  40. > where, for example, you still need morse code to get a ham license.

    except of course, you don’t. You can hold an extra class license without ever touching a key, or understanding the first iota of morse code.

    > To do UWB you need to convince the FCC (or substitute your local regulatory agency)
    http://www.fcc.gov/Bureaus/Engineering_Technology/News_Releases/2002/nret0203.html
    (I won’t forward the clearances in other countries.

    > spread spectrum and smart radios can allow you to tightly control what frequencies you use,
    spread spectrum, by definition is not ‘tight control of frequency’.

    > “junk” spectrum used by WiFi:
    I’m nearly certain that I started this meme, and its wrong.

    > Cell phones, for example, have managed to squeeze multiple orders of magnitude of
    > calls into the same badnwith over the twenty or so years they have been seriously around.

    Moore’s law marches on. Go take a look at the number of multiplies required for IFFT on your favorite
    OFDM-style modulation sometime.

    > The answer is of course there is, in the application we are talking about. If I want a room lit with a
    > soft sunlight glow, and you want the same room lit with a red glow, we are going to interfere, so in
    > that sense there is a scarcity of colors or at least in the use of colors which is what really matters.)

    Oooh boy. There is no interference of the colors (frequencies) in this (or, frankly any other) situation. You can certainly recover the spectra of interest via filtering and demodulation. (Think tinted glasses just before the receiver.)

  41. >I don’t know if there is a technical definition of scarcity (I am sure there is), but it seems to me there are two things that matter: if I am using it, can you use it too? (This would be my argment against using the word “property” in the context of intellectual property.) Secondly, is there a practical upper limit on the amount available for use? (This would be my argument against property rights in, for example, air.)

    Your intuition is sound. Economists capture the sense of your first question by asking whether a good is “rivalrous” or “non-rivalrous”. You and I agree that property rights are the best way to manage rivalrous goods. I, too, reject the idea that one should have to apply to a bureaucracy to use a radio — I no longer think property rights would be appropriate for radio spectrum, but I am clear that conventional regulation is worse.

    But your “property” analysis in the case of radio spectrum is in error. I sympathize with this error, since I used to make exactly the same mistake. You are correct that use of any given frequency band is a rivalrous good; that’s why, when we had to think of transmission in terms of fixed-frequency bands, property and homesteading was a good model. But when your radios are doing adaptive spread-spectrum, there’s no point in contesting frequency bands – if one is too noisy, you just jump to another.

    Thus, while individual frequencies are still rivalrous, total bandwidth is not. The upper limit on where you can jump to wouldn’t be scratched if you had a billion nodes on your network.

  42. I think you lack an understanding of spread spectrum technique.

    With spread spectrum radios, individual frequencies are not ‘rivalrous’, but the noise level in the occupied bandwidth *is*. Given this, the occupied bandwidth won’t survive 1 billion radios (assuming they’re all ‘active’.)

    Jumping to a new band is “adaptive”, not ”spread spectrum”. These are different things. Adaptive modulation (such as lowering the chip rate) is a different subject again, but here you’re just trying to lower the required Eb/No (SINR) enough for successful demodulation.

  43. @JessicaBoxer: you *do* realize that you just entered into a tautology? You pointed out that if I wanted to use one color in one room, and you wanted to use the same color, then there is a scarcity of that color. Well DUH! It’s trivially obvious that ONE color is scarce in the same location. My point, however, is that there are no end of colors we can use in the same room as long as we don’t all insist on using the same color, or there’s no end of reuse of the same color as long as we’re in different rooms. Efficient spectrum use means not transmitting any louder than you have to.

  44. @AndyFreeman says “The number of municipalities/co-ops that resist the temptation to filter will be dwarfed by the number who see filtering as one of the reasons to provide FTTH.” Hey, if the LDS don’t want to see any ads for strong drink come over their intarwebs, fine for them. Don’t live there. The rest of us will be rather insistent that the municipality or cooperative do nothing more than manage a stream of photons.

  45. @radu: by the way, the same thing happened in the railroad market. The railroads were suffering from ruinous competition. On their high volume lines, which should have been very profitable, they had to compete with other companies who saw those routes as very profitable as well. Look at http://russnelson.com/ny-rail-network.jpg (6K x 6K pixels; 3MB; may crash your browser). See the parallel lines between Erie and Buffalo, Rochester and Buffalo, Syracuse and Rochester, Albany and Syracuse, NYC and Albany, Watertown and Utica, Oswego and Syracuse, Port Jervis and Kingston. Everywhere else, you had high-cost, high-maintenance lines to regional markets.

    The ICC was created to fix railroad rates at profitable levels, and reduce competition.

  46. My point, however, is that there are no end of colors we can use in the same room as long as we don’t all insist on using the same color, or there’s no end of reuse of the same color as long as we’re in different rooms. Efficient spectrum use means not transmitting any louder than you have to.

    True, but its only a part of the solution. Another part is dense connectivity between nodes, and another part still is not employing OFDM (or worse, FM/AM modulation.)

    Of course, the bit rate for any one signal will go down. This is why its a mistake to talk about ‘bandwidth’.

  47. Eric: More recently, I have been convinced that frequency scarcity is an artifact of stupid radios. That is, if you have software-defined radios doing spread-spectrum packet protocols, it’s possible for two radios to negotiate a channel-hop sequence and get data through even under conditions of extremely high and time-varying noise from other transmissions.
    The problem with this is that going above a pretty small number of frequency hopping spread spectrum radios in a given band renders the band much less useful for existing users and modes of transmission. Can you say “flag day”? Cory Doctorow and I had an argument about this at the Penguicon he came to; I argued that using the ham bands for this experimentation on any scale would destroy ham radio as we knew it, and his reply was that it needed to be destroyed.

    Jessica: Things like UWB, spread spectrum, and smart radios in general, run into insane problems when the radio spectrum is managed by a creaking old bureacracy where, for example, you still need morse code to get a ham license.
    That hasn’t been true for any ham license since February 2007, and for at least one class of license since 1991.

  48. >Cory Doctorow and I had an argument about this at the Penguicon he came to; I argued that using the ham bands for this experimentation on any scale would destroy ham radio as we knew it, and his reply was that it needed to be destroyed.

    After some reflection, I’m with Cory on this one. The hams are supposed to be about experimentation with cutting-edge radio technology; it’s both their cultural raison-d’etre and the theory under which they have special status in the regulatorium (including the privilege of modifying transmission equipment). If they’re so attached to existing modulation techniques that they’re a drag on the move to adaptive smart radios, rather than leading it, then they’ve become a pointless old-boy’s club and should go to the wall.

  49. Ham radio has five purposes for existing, set forth in 47 CFR 97.1. Yes, experimentation is important, but it must be balanced with the other reasons for its existence.

    To take one example: Without ham radio, many agencies would have had no communications ability at all in the aftermath of Hurricane Katrina. The nature of radio means that folks using spread-spectrum techniques well outside the disaster area nevertheless would affect disaster communications within the area. By its nature, spread spectrum users would not know that they are causing harmful interference.

    Don’t get me wrong: I’d love to see hams do that work. I just don’t think it should be pursued to the exclusion of everything else that ham radio is, and that’s what would wind up happening unless great care is taken to balance out the effects.

  50. Looking at 47 CFR 97.1 with this issue in mind:

    >(a) Recognition and enhancement of the value of the amateur service to the public as a voluntary noncommercial communication service, particularly with respect to providing emergency communications.

    OK, that’s the don’t-disrupt-emergency-services argument.

    >(b) Continuation and extension of the amateur’s proven ability to contribute to the advancement of the radio art.
    >(c) Encouragement and improvement of the amateur service through rules which provide for advancing skills in both the communications and technical phases of the art.
    > (d) Expansion of the existing reservoir within the amateur radio service of trained operators, technicians, and electronics experts.

    All three of those weigh on Cory’s side. It would be perverse if the statute were interpreted as a mandate for preserving obsolete techniques and skills.

    >(e) Continuation and extension of the amateur’s unique ability to enhance international goodwill.

    That one seems orthogonal to the issue.

  51. In regards to color sharing, and the problem of using unlicensed spectrum: I am totally on board with the idea that new technologies allow much more efficient use of the radio spectrum; however, Claude Shannon is not sweating it. There is an upper limit to the amount of data you can pack into the bandwidth available, and there are only a restricted range of frequencies that are usable. That makes it a resource with limited capacity, and the only efficient way I know of to utilize that is private property.

    In the days of the old west, cattle were grazed on common (government owned) ground, and the classic commons problem did not immediately appear. Why? There were only a limited number of ranches grazing the ground. But as time went by, more and more farmers moved in. Some didn’t follow the established standards (for example thouse cotton pickin’ sheep farmers, whose sheep alegedly cropped the grass too short), and eventally the commons had to be fenced off, and allocated privately (or managed by a government bureacracy.)

    My point is that new technology might open up the commons to vastly increase the amount of data throughput. If you are nowhere near the limits, the commons works OK. That is where we are at. However, data throughput is an insatiable monster, we will use more and more of it, and eventually, be bumping up against the limits. Better to fence it off now before the ranchers start shooting the shepherds.

  52. I would argue that 47 CFR 97.1(c) and (d) weigh in equally on both sides, since they’re not solely about technical advancement: they’re also about training people to communicate and operate effectively. Unless we perform a forklift upgrade on every piece of communication technology currently in use, there’s still plenty of need to train people on the stuff we’re using now.

    I’ll agree that (e) is mostly orthogonal to the issue, but it does mean that that forklift upgrade needs to be worldwide…

  53. Looking at 47 CFR 97.1 with this issue in mind:

    >(a) Recognition and enhancement of the value of the amateur service to the public as a voluntary noncommercial communication service, particularly with respect to providing emergency communications.

    OK, that’s the don’t-disrupt-emergency-services argument.

    >(b) Continuation and extension of the amateur’s proven ability to contribute to the advancement of the radio art.
    >(c) Encouragement and improvement of the amateur service through rules which provide for advancing skills in both the communications and technical phases of the art.
    > (d) Expansion of the existing reservoir within the amateur radio service of trained operators, technicians, and electronics experts.

    All three of those weigh on Cory’s side. It would be perverse if the statute were interpreted as a mandate for preserving obsolete techniques and skills.

    Never mind that HAMs were doing spread spectrum radio in the 80s, going so far as to publish the “ARRL Spread Spectrum Sourcebook”.

    Never mind that HAMs are living on the cutting edge of software-defined radios.

    Never mind that HAMs are on the cutting edge of digital techniques, and have been doing these for years, while FM radio continues on…

  54. Oh, Jim, I agree (as I look at the stack of D-Star radios next to the monitor on my desk). I’ve got no problem at all with spread spectrum experimentation. Where Cory and I part company is that he called for, essentially, spread spectrum experimentation on all bands, with no regard for existing uses. That will never work in the real world, if for no other reason than any attempt to amend Part 97 to allow it will result in a firestorm of comments from hams about how it’s a Bad Idea.

    One quibble: “ham” is not an acronym. The origins of the term are unclear, but all agree that it’s properly treated as a noun, and uncapitalized.

  55. I’m willing to support SS operation in a portion of nearly all bands. Then only real issue is occupied bandwidth.

    Frankly, for ham radio to survive, the FCC has to allow this, because the Elmers are dying off, and the young crowd has the Internet to chat.

    Ham radio has to re-invent itself. Digital modes are one such way.

  56. Some thing really interesting no notice: ESR who believes Wikipedia is a disaster makes two references to it’s articles here(Mesh networking and regulatory capture).What do we call a person who says one thing and does another?!

  57. >ESR who believes Wikipedia is a disaster

    I’ve never said that. Wikipedia is pretty good for facts; it’s poor where editorial judgment is required, and abysmally poor on topics that are politically loaded.

  58. >I’ve never said that.

    “…he was particularly critical of Wikipedia, calling it a “disaster”,…”
    I extracted this part from the Wikipedia article here : http://en.wikipedia.org/wiki/Eric_S_Raymond
    which makes a refence to this link as a source :
    http://www.newyorker.com/archive/2006/07/31/060731fa_fact?currentPage=5

    >it’s poor where editorial judgment is required, and abysmally poor on topics that are politically loaded.

    Anyhow,your argument sounds reasonable and I believe it too.I just thought that Wikipedia as a whole is really a good project and when you yourself accept that I guess we are in the same boat here.

  59. Net neutrality must come via natural selection (an economic model), and I think P2P at the protocol layer (i.e. protocols on top of TCP/IP, UDP) may be the best candidate, because it can be faster and more reliable than the (undistributed, centralized manager) server/client model, so once all content is served distributed, then who can the last mile players send a bill to, other than directly to the end user? People can be tricked by silent taxes, such as inflation, but they usually seek out competition when the bill has their name on it. This then renders 666 (fascist money), patents, copyrights impotent if it is anonymous (e.g. MUTE, but with some necessary improvements I have recently in mind, not yet documented publicly), which will of course remove Linux from it’s media codec conundrum (and a zillion other magnificient prosperity effects for mankind, e.g. making gold trading anonymous, etc).

    http://forum.bittorrent.org/viewtopic.php?pid=166#p166
    http://ncannasse.fr/blog/captcha?lang=en#comments (find my comment)
    http://lists.motion-twin.com/pipermail/haxe/2008-November/020935.html

  60. I agree with Jim that open fiber networks are the real solution here, every thing else is chaff. Former FCC economist Tom Hazlett notes that the white spaces crowd should be going after the bandwidth set aside for digital TV, 90+% of which isn’t being used, rather than trying to fit into the empty spaces in between (source – http://www.econtalk.org/archives/2008/11/hazlett_on_tele.html). Similarly, the NN crowd shouldn’t be trying to put these puny restrictions on the telcos, they should be joining forces with us to build a whole new open fiber network to the home, where the COs are open to all ISPs that want to deploy hardware and the local fiber loop is unbundled. The fact that both the white spaces and NN crowds are too stupid to realize this is a testament to their irrelevance. I’m confident that someone will build an open fiber network someday, simply because it’s too obvious not to be done. They seem to have tried something like it in Utah – http://en.wikipedia.org/wiki/Utah_Telecommunication_Open_Infrastructure_Agency

    esr, you claim monopolies only come about from regulation, how do you explain almost-monopolies like MicroSoft or Intel then? Surely there’s room for a company to so dominate its field that it becomes pretty much a monopoly. I actually think there’s another reason for monopolies to form that’s not often talked about: idiot investors. People like Warren Buffet want to invest in monopolies, in his case in the newspaper business, so that their returns are more assured, not realizing or caring that this slows down the process of competition and innovation, that we all benefit from. I think similar reasoning leads many to invest in monopolies like MicroSoft or cheer on rampant mergers till there is only an AT&T and Verizon left- not because the resultant mega-corps are more efficient, which is the superficial excuse- but simply so investors can extract as much monopoly rent as they can (another possible reason is that investors are lazy and it’s easier to choose between two mega-corps to invest in than many smaller businesses). Also, egotistical managers would much rather boss around 100k employes than 1k, given the chance. Dumb arguments, like economies of scale, have been manufactured to support this market pressure for monopolistic corporations. What’s funny is that these monopolies always splinter and fail soon enough, when like all bureaucracies they’re eventually unable to accomplish anything under the anchor of their own weight. This is what is happening today as the dimwits in the newspaper industry face competition for the first time, from online advertising, and as Microsoft flies off into irrelevance as they copy their dumber competitors like Sun or Adobe, because they don’t have a more successful competitor to copy, which has always been their M.O (not that there’s anything wrong with copying, it’s just ridiculous given all their talk about supposed innovation).

  61. >esr, you claim monopolies only come about from regulation, how do you explain almost-monopolies like MicroSoft or Intel then?

    Um, no. What I believe, and have written, is that monopolies have a short half-life in the absence of government market-rigging – they arise spontaneously in free markets, but they also pop like soap bubbles. This half-life has been measured by economic historians; it’s about twelve years. You’ve noted some of the reasons they’re unstable in the comment I’m replying to.

  62. OK, thanks for the clarification, I think we’re agreed then. In fact, great original post laying out the current telco situation, mirroring my own thoughts almost exactly, with the only exception that I think open fiber networks are the route around this mess rather than any wireless scheme. All the wireless schemes require fiber injection-points to do the real heavy lifting anyway, so fiber is the real key here. It is a shame that the current regulators do not enable open fiber networks but I see no reason they can’t just be ignored and pushed aside anyway.

  63. No you are missing the point, that the physical means of networking (even wireless mess) can always be controlled by society and thus victim to govt enforced monopolies.

    Go back and re-read my prior post. The way forward is an economic solution, whereby through anonymity of data server (and even encryption between end client and server), then the crooks can not charge any one but their own customers directly. The problem we are trying to solve is one where the crooks are (threatening to) filtering content based on the identification of it’s source, in order to make the source (e.g. Google) pay. If the sources can not be charged, then the socialism ad-funded model of the internet can also die, before it overcentralizes the economics of the internet and strangles it.

  64. To elaborate on your last comment ” Oh, and buy an Android phone. As I noted in my immediately previous post, Google is our ally in this.” I did by a Android phone last week and I absolutely love it. It completely blows the Iphone or another smart phone out of the water. Long Live Google !!

  65. >Rand was a brilliant moral critic but an embarassingly, cringe-makingly bad epistemologist. I might blog a detailed takedown of her epistemology sometime;

    Did you ever blog on this subject? If so, could you post a link to it? Thanks.

  66. On the off-chance you’ll read this, the link to your discussion on “the last mile” appears to go right back to this same blog post.

  67. You are truly ignorant. “The network neutrality crowd are a bunch of leftist”. What a stupid assumption I am not a liberal, nor a libertarian nor a conservative. I am simply guided by the reality of telcos and governments, they are cuddled together and willing to screw people over to make the internet more like a cable service. You are ridiculous

  68. All the NN discussion I’ve seen has focussed almost exclusively on filtering and QoS, but I think that’s really a side issue, and the real linchpin of this is the economics of peering arrangements

  69. All the NN discussion I’ve seen has focussed almost exclusively on filtering and QoS, but I think that’s really a side issue, and the real linchpin of this is the economics of peering arrangements

    No. When I go onto YouTube with my Comcast connection, pings to 8.8.8.8 drop from 35-50ms to 3000-5000ms — meaning that when Comcast sees packets coming from YouTube, they throttle me by a factor of about 100!

    This is not a peering arrangement problem. This is a “Comcast deliberately abusing its middleman position to collect rent from me and/or Google” problem. And it is illegal.

    In Europe, anyway. Here in Murka the rule seems to be “what’s good for the megacorps is good for society”.

    This is why we need net neutrality regulation.

Leave a comment

Your email address will not be published. Required fields are marked *