Zero doesn’t apply to spectrum

by on May 17, 2006 · 6 comments

Mike at TechDirt takes issue with an article by Tom Lenard on C|Net that argues that market allocation of “white space” spectrum is more efficient than a “commons” designation. He writes that “Unlicensed spectrum is hardly a ‘centralized allocation system,’ and it’s hard to see how anyone could make such a claim with a straight face.” As I explained in a recent paper, in order to have a “commons” that works, you need to have rules that govern how devices operate in the space so that they don’t interfere with each other. For example, devices in the chunk of spectrum in which Wi-Fi operates, by regulation, cannot operate above 5 Watts EIRP. Therefore, the rules that govern the “commons” we now have are centrally planned by the government. It’s not controversial to say that central planning is inefficient because a planner cannot possibly have all the information about all the possible competing uses of the spectrum.

While there no doubt is a place for unlicensed devices, one has to admit that designating some spectrum as a commons with certain specific rules will prevent that spectrum from being used in another, perhaps more innovative way, that cannot operate within the commons’ rules. In a market you could just buy the spectrum and deploy your more innovative use; in a commons regime you would have to petition the central planner to change the rules (and we all know how well and how quickly that works.)

Second, Mike takes Lenard to task for not understanding the concept of zero, claiming that when scarcity is removed, market oriented folks have a hard time understanding policy. I tend to agree with him, and I think his is a great observation as it applies to intellectual property. Ideas truly are not scarce; their scarcity is created artificially through IP laws. However, I’m afraid that while new technologies have been able to eke out more communications capacity from existing spectrum, that capacity is still finite and, despite the rhetoric one often hears, spectrum scarcity has not been eliminated.

Mike writes that “what those who understand zero recognize, is that unlicensed spectrum turns spectrum into a free input, lowering the costs and allowing companies to provide products that serve the market at much more reasonable rates.” What he doesn’t see is that while unlicensed spectrum might be a “free input” for certain uses, a whole host of other uses are precluded. While a commons can allow low power, near range devices such as Wi-Fi, bluetooth, and cordless phones–great innovations all–you could not deploy a new national wireless competitor in voice or video over unlicensed spectrum. The only way the cost of spectrum could truly be zero is if all potential uses of spectrum could be deployed without precluding any other use. This is the case in intellectual property where I can use any idea as much as I want without ever affecting someone else’s ability to use that same idea. But it’s not the case for spectrum where one use of spectrum (even the use of spectrum for an unlicensed commons) will necessarily preclude some other potential use.

  • http://www.techdirt.com/ Mike Masnick

    Hi Jerry,

    I responded to your comment directly on Techdirt, but I think this is a bit trickier than maybe either of us describe. I’ll agree that where interference is an issue, then you have a non-zero situation, but in many cases with spectrum you can have a situation that is effectively zero — which is the case with WiFi — though, admittedly that involved some borderline FCC rules. However, to jump from those rules to “central planning” is a big stretch. If you notice what those rules were designed to do, they were specifically designed to mostly eliminate the problem of interference (i.e., pushing it back to a “zero” situation). And it worked. It would be difficult to argue that the open spectrum didn’t stimulate an awful lot of innovation.

    I agree that the gov’t won’t always be able to set the rules so well, but… there’s pretty clear proof that you can set things up in a way where open spectrum does encourage innovation.

    In the case of wide area deployments, where interference may be a bit more of a problem, you have a slightly different situation. The interference problem hurts everyone — it doesn’t benefit one at the expense of others. In that case, it behooves everyone who has a stake to set the rules themselves so that interference isn’t a problem. So, the case here is a bit different, but because you have an all or nothing situation, zero makes a reappearance in the idea that in the situation where there’s interference, the entire supply drops to zero.

    So, I’m all for licensed spectrum, and giving that to certain providers who have incentive to make the best of it. However, there is also clear value in opening up spectrum with very basic guidelines of usage and letting competition happen within the spectrum.

    Those basic rules are guidelines to minimize the problem of interference, not top down central planning, and I think it’s a stretch to jump from one to the other.

  • http://www.techliberation.com/ Tim

    Mike:

    If we had a regime in which all spectrum was privately owned, and owners had the right to do whatever they wanted with it (as long as their use didn’t interfere with other bands), wouldn’t it be in the interest of some private companies to set up their spectrum as a commons themselves? You could charge, say, $1 per device to any manufacturer that wanted to use the spectrum. If it’s true that setting up some spectrum as a commons is more efficient than setting it up all up for exclusive use, then companies that set up spectrum commonses would make more money than companies that tried to use it only for themselves.

    Moreover, even if we agree that having a spectrum commons is a good idea (which I do) it’s not obvious what the best rules are. Should devices be limited to 1 watt, 5 watts, or 25 watts? Are there smart protocols that would allow devices to negotiate spectrum sharing on the fly? Are there ways to design devices to minimize interference? The list can go on.

    Obviously, those kinds of decisions are much less intrusive than having the government allocate exclusive licenses that specify who’s going to use the spectrum and how. So a government-managed commons will work better than the command-and-control scheme. But it still seems to me that privately-run commons are likely to work better, because the company has an incentive to choose rules that maximize the use (and therefore the revenue) of the commons.

    Moreover, there’s an important question concerning how much of the spectrum should be allocated to proprietary versus common use. The FCC doesn’t have any real basis for making that decision, because without some way of pricing the competing uses, theres no way to determine the economic value of the common uses. In contrast, private companies would have an incentive to convert exclusive spectrum to common spectrum (or vice versa) until the revenue generated by each is equal. Which is, I think, the efficient outcome.

    The real problem, I think, is that the FCC doesn’t allow private companies to do what they want with spectrum they license. The license typically specifies not only who gets to use the spectrum, but what they should use it for as well. Liberalizing those rules would allow spectrum to migrate to its highest-valued use without the need for omnicience on the part of the FCC.

  • http://www.techdirt.com/ Mike Masnick

    Hi Jerry,

    I responded to your comment directly on Techdirt, but I think this is a bit trickier than maybe either of us describe. I’ll agree that where interference is an issue, then you have a non-zero situation, but in many cases with spectrum you can have a situation that is effectively zero — which is the case with WiFi — though, admittedly that involved some borderline FCC rules. However, to jump from those rules to “central planning” is a big stretch. If you notice what those rules were designed to do, they were specifically designed to mostly eliminate the problem of interference (i.e., pushing it back to a “zero” situation). And it worked. It would be difficult to argue that the open spectrum didn’t stimulate an awful lot of innovation.

    I agree that the gov’t won’t always be able to set the rules so well, but… there’s pretty clear proof that you can set things up in a way where open spectrum does encourage innovation.

    In the case of wide area deployments, where interference may be a bit more of a problem, you have a slightly different situation. The interference problem hurts everyone — it doesn’t benefit one at the expense of others. In that case, it behooves everyone who has a stake to set the rules themselves so that interference isn’t a problem. So, the case here is a bit different, but because you have an all or nothing situation, zero makes a reappearance in the idea that in the situation where there’s interference, the entire supply drops to zero.

    So, I’m all for licensed spectrum, and giving that to certain providers who have incentive to make the best of it. However, there is also clear value in opening up spectrum with very basic guidelines of usage and letting competition happen within the spectrum.

    Those basic rules are guidelines to minimize the problem of interference, not top down central planning, and I think it’s a stretch to jump from one to the other.

  • http://www.techliberation.com/ Tim

    Mike:

    If we had a regime in which all spectrum was privately owned, and owners had the right to do whatever they wanted with it (as long as their use didn’t interfere with other bands), wouldn’t it be in the interest of some private companies to set up their spectrum as a commons themselves? You could charge, say, $1 per device to any manufacturer that wanted to use the spectrum. If it’s true that setting up some spectrum as a commons is more efficient than setting it up all up for exclusive use, then companies that set up spectrum commonses would make more money than companies that tried to use it only for themselves.

    Moreover, even if we agree that having a spectrum commons is a good idea (which I do) it’s not obvious what the best rules are. Should devices be limited to 1 watt, 5 watts, or 25 watts? Are there smart protocols that would allow devices to negotiate spectrum sharing on the fly? Are there ways to design devices to minimize interference? The list can go on.

    Obviously, those kinds of decisions are much less intrusive than having the government allocate exclusive licenses that specify who’s going to use the spectrum and how. So a government-managed commons will work better than the command-and-control scheme. But it still seems to me that privately-run commons are likely to work better, because the company has an incentive to choose rules that maximize the use (and therefore the revenue) of the commons.

    Moreover, there’s an important question concerning how much of the spectrum should be allocated to proprietary versus common use. The FCC doesn’t have any real basis for making that decision, because without some way of pricing the competing uses, theres no way to determine the economic value of the common uses. In contrast, private companies would have an incentive to convert exclusive spectrum to common spectrum (or vice versa) until the revenue generated by each is equal. Which is, I think, the efficient outcome.

    The real problem, I think, is that the FCC doesn’t allow private companies to do what they want with spectrum they license. The license typically specifies not only who gets to use the spectrum, but what they should use it for as well. Liberalizing those rules would allow spectrum to migrate to its highest-valued use without the need for omnicience on the part of the FCC.

  • http://www.techdirt.com/ Mike Masnick

    Tim,

    The real problem, I think, is that the FCC doesn’t allow private companies to do what they want with spectrum they license. The license typically specifies not only who gets to use the spectrum, but what they should use it for as well. Liberalizing those rules would allow spectrum to migrate to its highest-valued use without the need for omnicience on the part of the FCC.

    That, I believe, hits the nail on the head. I absolutely agree with the above, and have said so repeatedly. That’s why I’m encouraged by what the UK is doing with spectrum — effectively letting the owners do what they want with it.

    However, back to the commons question… I actually think that it may be a similar situation as the one you pointed to in another post about standards with many patent holders involved. With that many stakeholders, a balance is reached that keeps any one player from ruining it for everyone…

  • http://www.techdirt.com/ Mike Masnick

    Tim,

    The real problem, I think, is that the FCC doesn’t allow private companies to do what they want with spectrum they license. The license typically specifies not only who gets to use the spectrum, but what they should use it for as well. Liberalizing those rules would allow spectrum to migrate to its highest-valued use without the need for omnicience on the part of the FCC.

    That, I believe, hits the nail on the head. I absolutely agree with the above, and have said so repeatedly. That’s why I’m encouraged by what the UK is doing with spectrum — effectively letting the owners do what they want with it.

    However, back to the commons question… I actually think that it may be a similar situation as the one you pointed to in another post about standards with many patent holders involved. With that many stakeholders, a balance is reached that keeps any one player from ruining it for everyone…

Previous post:

Next post: