So I’m an extremist

by on September 9, 2010 · 3 comments

After our podcast last week, Tim Lee wrote a blog post expanding on our conversation about spectrum policy. I thought I’d take a little space here to respond.

Although we will probably continue to disagree on empirical questions, I think philosophically there is no light between Tim and me. He succinctly expresses our shared view when he writes,

The question advocates of free markets (“extreme” or otherwise) need to ask is not: which property rights should we create? Rather, we need to ask: which set of regulations prevents congestion at the lowest cost in liberty? You want the set of rules that maximizes individual freedom; rules that are clear and predictable and give government officials as little opportunity as possible to make mischief.

Tim and I simply come to different conclusions in our cost-benefit calculation when we look at the competing sets of rules for spectrum.

As Tim points out, radio spectrum is not like sunlight. We can’t use as much of it as we want without congestion. This precludes an open access regime. The question then is, do we create a regime were private actors own the spectrum and make decisions about how to best utilize it? (Government’s only function in this scenario would be to enforce contracts and property rights.) Or do we allow the government to in essence own the spectrum and determine which specific uses will be permitted? (Government in this scenario decides the rules that govern a commons, which is not a trivial matter since it will allow some uses and preclude others.)

So, which regime presents “a lower cost in liberty?” Which one is more “clear and predictable?” Which one presents “government officials as little opportunity as possible to make mischief?” I tend to think the former fits the bill much better than the latter. Tim thinks we can have a little bit of both; that both can be equally efficient. He seems to see little difference between government-as-court and government-as-regulator. I’m less sanguine about government’s ability to set rules for spectrum that get you to the same or better outcome than a property rights regime.

In case there’s any confusion about it, I agree with Tim that free-market principles do not demand that everything be propertized. But in the particular case of spectrum, I believe property and markets are more efficient than either command-and-control allocation or a government commons. And yes, I’m talking about the whole spectrum. Tim identifies the relevant questions to ask in making this determination:

[W]hich set of regulations maximizes the freedom of individuals to use the spectrum as they choose? And which set of regulations will lead to the most efficient utilization of spectrum?

I think the second question informs the first. Without some form of resource allocation, a tragedy of the commons ensues and no one gets to use the spectrum. So, what we need to find is the set of rules that get us the most efficient and most valued uses.

Tim says that my “preferred scheme of exclusive licenses for the entire spectrum doesn’t fit the bill because it puts a thumb on the scale in favor of large, capital-intensive firms that can win multi-billion dollar auctions.” First, I don’t see why this should be the case. The reason why spectrum auctions today fetch billions of dollars in revenue is that the vast majority of the radio spectrum is either in civilian government or military hands, or in private hands but not tradable and limited by law to specific uses. Only a small sliver of it has been auctioned as flexible-use licensed spectrum. In a world of privatized radio frequencies, the auction price of spectrum would fall relative to what we see today.

But putting that aside, how are capital-intensive uses necessarily inconsistent with the most efficient utilization of spectrum? Tim seems to be suggesting that even if people highly value “small-scale, short-range applications like WiFi,” there is no way that a market could provision these. I don’t see why that should be the case. Tim pooh-poohs the notion that a firm (maybe a consortium like the Wi-Fi Alliance?) or a non-profit spectrum conservancy as I have been suggesting could adequately provide the use. He says that such a scheme would be relegated to second-class citizen status, but he doesn’t explain why.

Tim says that what’s needed is a policy that accommodates both long-distance high-powered uses like broadband or wireless backhaul, and short-range low-power commons-type uses like Wi-Fi and Bluetooth. The sort of property- and market-based policy I have in mind is exactly that.

Tim also says that a “decision that all spectrum will be exclusively licensed is a government choice whose consequences can’t be fully corrected by subsequent market transactions.” First, what are these problematic consequences that will need correction? Second, if correcting them means that government needs to have control over some of the spectrum, it always has the power of eminent domain. In a spectrum market setting, we would have a much better sense of the true opportunity cost of placing spectrum under government management. Because it would have to pay just compensation for what it takes, the government would have to internalize the true costs of its decisions. And of course government can avoid using a takings. I can imagine a big-bang spectrum auction where the government is one of the bidders for some of the spectrum—again, fully internalizing the true cost of the spectrum.

Finally, I hope it’s clear why I’m so skeptical of government’s ability to manage a spectrum efficiently. It’s for all the reasons Tim masterfully lays out when he talks about bottom-up solutions generally being superior to top-down ones. Government planners do not have the information or incentives to set the most efficient rules. Worse, it’s likely that the rule-setting process is corruptible so that one possible set of rules is favored over another for political reasons. And talk about correcting unintended consequences. I’m pretty sure Tim will agree markets tend to be more speedy than bureaucracies.

  • http://openid-provider.appspot.com/quanticle quanticle

    And talk about correcting unintended consequences. I’m pretty sure Tim will agree markets tend to be more speedy than bureaucracies.

    Only if the market is free. Right now, I dread calling Comcast's customer service far far more than I dread going to the DMV (actually, the local DMV isn't bad at all). I don't think its necessarily self evident that private “market driven” firms are any better at allocating capital and minimizing bureaucracy than the government.

    We libertarians need to realize that markets != free market. A monopoly or an oligopoly is a market too.

  • http://twitter.com/binarybits Timothy Lee

    Thanks for a great response Jerry! I think you accurately summarized my views and made our areas of disagreement clear. So let me just address this:

    Tim seems to be suggesting that even if people highly value “small-scale, short-range applications like WiFi,” there is no way that a market could provision these. I don’t see why that should be the case. Tim pooh-poohs the notion that a firm (maybe a consortium like the Wi-Fi Alliance?) or a non-profit spectrum conservancy as I have been suggesting could adequately provide the use. He says that such a scheme would be relegated to second-class citizen status, but he doesn’t explain why.

    Transaction costs matter. AT&T already has the infrastructure and management hierarchy to run a nationwide wireless network, so a nationwide spectrum license fits its needs perfectly.

    In contrast, imagine a small company in BritoLand that wants to use WiFi in its own office. To get permission to do so, it's going to have to band together with other small companies to organize a nationwide bureaucracy whose only purpose is to identify potential users of small bits of spectrum, collect money from them, use that money to win an auction, and then disband afterwards. Since this bureaucracy wouldn't otherwise exist, its creation is a pure deadweight loss. Is it possible that market actors would organize themselves in this way?

    Would market actors figure out how to do this? Sure, if the potential value were high enough, they probably would. But it strikes me as a perverse way to go about it. The asymmetry in organizational transaction costs (the fact that AT&T is already a bureaucracy but the WiFi alliance has to create one from scratch) means that the scheme will systematically over-allocate spectrum to the AT&Ts of the world and under-allocate it to commons-style uses.

    Imagine an alternative scheme in which everyone gets a property right that allows them to broadcast on any frequency but only within their own property lines. And suppose the scheme allowed you to re-assign this right on a frequency-by-frequency basis (so I can sell the right to transmit at 700 MHz on my property to whomever I want). Notice that this scheme allocates spectrum rights purely according to market forces.

    It's theoretically possible that AT&T could go door-to-door buying up everyone's spectrum rights, and thereby assemble a national spectrum license. But of course this would be a fantastically expensive process. In the real world, this system would give us a lot of WiFi-style broadcasters and very few long-range ones.

    The two cases are mirror images. If it's ridiculous to expect AT&T to assemble tens of millions of spectrum rights in order to broadcast nationwide, it's equally ridiculous to expect me to purchase (part of) a nationwide spectrum license to get permission to broadcast within the privacy of my own home. Choosing either extreme—all exclusive or all unlicensed—imposes unreasonable burdens on those who would have preferred the opposite regime. The fact that markets are theoretically capable of overcoming this distortion doesn't mean it's fair or efficient to make them do so.

  • http://pulse.yahoo.com/_IBLTVRF2367QCEAUKSVVW6CVR4 Anonymous Mouth

    I think that proponents of property rights in spectrum should be obliged to deal with the issues raised by Hatfield and Weiser in numerous papers (e.g. http://www.cato.org/pubs/pas/pa575.pdf). I am not confident that even specialized courts can ever be competent to solve all potential “trespasses” in spectrum–especially because, in spectrum, the standard for “harmful interference” (the only meaningful measure of interference–zero spillover is physically impossible) should change with technology. Nor is it clear to me how a new “easement” for a service like ultra wideband could develop through common law principles.

    It's not enough to argue for the efficiency of property rights in general–the specifics of propagation and spillover must be addressed, in advance.

    Jerry, while I appreciate your candor in advocating against even successful unlicensed commons approaches (your “opportunity cost” discussion in the podcast), I think the problem I have both with property rights “extremists” and commons extremists, is that the discussion is all very high-level political and economic theory. I think the world of spectrum is too messy and complicated for any one system to work, and we have had success with unlicensed and licensed-with-strings attached approaches. I don't want to give up those portions of our current system until I see a property rights system actually working. I guess my position is “spectrum conservatism.”

Previous post:

Next post: