It would be great help if I could get some clarity on the real issues…

After all, there was a time when wages for certain professions were regulated across whole kingdoms. It wasn’t fatal. Life went on, but that doesn’t justify the regulation.

Treating this as an ideological issue, simply declaring that “the market is always right”, is too simple-minded for such a complex issue.

Have I ever said such a thing? My position would be more accurately described as “The market is so beneficial on the whole that I don’t support intervention unless a certain inefficiency is simply unbearable.”

You could convince me that such is the case, but you haven’t made much progress yet. That’s all I’m saying.

The question is not whether we should let politics determine these issues. The question is whether we should have a central referee who sorts out the issues and imposes some standards on a very complex situation. There are unquestionably some major economic benefits that we enjoy due to having this central referee. There are also some economic losses we pay for it. Treating this as an ideological issue, simply declaring that “the market is always right”, is too simple-minded for such a complex issue. I’ve been dumping kilobytes of text on you guys trying to lay out just SOME of the complexities. I don’t demand that you accept my personal conclusions. But you would be foolish to ignore the complexities I’m explaining.

Let’s just concentrate on those situations in which state intrusion results in better performance due to technical considerations.

Actually, I’d rather stick to electromagnetic communications. I take issue with a number of your assertions in the other examples as well, so the discussion would quickly get out of hand.

This delays adoption of a beneficial technology. If the state had simply declared “We’re going to use Technology A”, then we’d have earlier adoption of the technology.

You’re making the unjustified assumption that both paths would begin at the same time. It’s not a stretch to suggest that the market may be able to go through several iterations of “beta” products before the government has even recognized the opportunity, let alone reached a consensus on a path.

the basic physical layouts are the same and the only differences are in data structures, so format issues aren’t so serious in that case.

Ok, but I have a $20 media reader that can read and write 7 types of solid state media. Let’s try to avoid tangents.

Anyway, this process of setting technical standards is really hairy and just intoning “let the market do it” doesn’t address the complexities.

Indeed, but your assertion is that adding electoral politics to the mix is going to improve things enough to justify the intervention.

I remain unconvinced.

One example might be roads. Yes, toll roads can work in some special cases, but for the basic road net, a privatized system would be most inefficient. Having to stop every few blocks to pay a 3 cent toll here and a 8 cent toll there is just not a good way to get around. Yet if we rely on economies of scale, we get into problems of monopolistic behavior — when there’s a big event at the theater, the owner of the roads jacks up the toll. This is no way to run a road system. And we all agree that this is a case for state ownership, not private ownership.

The same thing goes for public utilities such as fire and police protection, electricity, and telephone wires. So the real question is whether we should treat radio communications as a public utility. It’s a complicated issue, as I’ve been pointing out.

In general, the ones that lose out in cases like you’ve cited are “early adopters” who readily accept that risk as part of the transaction.

Which means that rational consumers will wait until long after the technology is ready, waiting to see which format will win out in the end. This delays adoption of a beneficial technology. If the state had simply declared “We’re going to use Technology A”, then we’d have earlier adoption of the technology. Of course, we all understand that the state could easily make the wrong choice (and often does). My point is that there is some economic benefit to having a central referee flip a coin and make a decision when the call is very close.

For every BetaMax example, there are many more where the free market faced such a choice and decides “I want both standards!” Witness optical disc drives that can read and write to an incredible combination of disc formats and speeds.

Well, yes and no. It depends on how truly different the standards are. With optical disk drives, the formatting differences are really tiny; the basic physical layouts are the same and the only differences are in data structures, so format issues aren’t so serious in that case.

Let’s take an example of a situation where differing standards have been worked out by the market successfully. Let’s talk about communications buses. We go all the way back to the venerable (and terribly confusing) RS-232 standard, which is still in use in some older equipment. (I have a few devices that still use it.) It was in use in the 70s. It’s a pretty good serial standard, but it is slow and stupid; you can’t share the bus with any other devices, so you have to dedicate a separate port to each device. Through the years we’ve seen all sorts of buses: the Centronics printer bus, SCSI, ADB, USB, FireWire, USB2, and so on. The market has done a good job of maintaining enough diversity to permit technological advances while providing some standards and stability for consumers. But there’s a key reason why: there is only one group of companies involved in creating the standards: the hardware manufacturers. And even then the standardization process is pretty simple: most new designs are readily presented to the relevant industry standards committee to obtain registry, so that they can license the standard to the commodity hardware manufacturers. Here, everybody’s interests are aligned and it’s easy to get a standard.

But with telecommunications, it’s immensely more difficult, for many reasons. In the first place, there are many more parties: the people who manufacture the transmission equipment, the people who manufacture the reception equipment, the people who provide the service, and the content people. Getting all these people to agree on anything can be hellacious. I remember sitting in on a single standards committee meeting as a content expert way back in 1984, and the discussion was hopelessly muddled. The guys who were running it (Philips) were trying to get a standard for CD-ROM reading devices. The hardware people wanted an absolutely minimal device that would just read data and send it down a bus like a hard drive. But CD is much slower than a hard drive, and as a content provider, I felt that we needed a lot more intelligence to make it useful. I remember waiting for an hour to get my two cents’ worth in. Finally, when my turn came, I screwed up my courage and said, “I think we need to put in at least a 68000 CPU to make this thing run fast enough.” There was stunned silence throughout the room; my proposal was really out on the far edge. The silence was broken by Alan Kay, one of the great computer scientists of our age, who said, “Is that all?” as if a 68000 were too weak a CPU for this. I was absolutely flabbergasted, and I cannot imagine what the Philips guys thought of Alan Kay. In any event, the CD-ROM standard for readers did not include a 68000, and in fact was pretty dumb, which was one reason why CD-ROMs took a long time to catch on. They were just too slow.

Anyway, this process of setting technical standards is really hairy and just intoning “let the market do it” doesn’t address the complexities. This is not like selling potato chips.

Note, however, that while combining deeds creates wealth, breaking them down destroys wealth. This suggests an inexorable market bias towards combining deeds — which could get us into monopoly issues.

Only to a point. The optimum deed size will probably be one that covers most of a city. After that, the returns would diminish quickly. Monopoly issues would be minor because you’d still have other channels.

When the dust settled, you’d probably have something remarkably similar to the status quo without having to take a top-down approach.

Two other points:
- In general, the ones that lose out in cases like you’ve cited are “early adopters” who readily accept that risk as part of the transaction. (I don’t regret buying an LS-120)
- For every BetaMax example, there are many more where the free market faced such a choice and decides “I want both standards!” Witness optical disc drives that can read and write to an incredible combination of disc formats and speeds.

Now, I’m not suggesting that we need to Federal government to dictate that everybody dump Windows and embrace Mac or Linux. My point is that, in some technical markets, the free marketplace does NOT yield optimal results. Blind faith in the efficiency of the marketplace in such situations is misplaced.

A variation of this phenomenon is called “winner takes all”. In some kinds of markets, the market leader simply wipes out all competition. A good example of this is the Windows/Mac competition, or the Netscape/Explorer competition. Google is presently using its market size to crush competition, and again, some of its advantage stems not from a superior product but from greater size. The end result in each case is a failure of the market to deliver the optimal results.

The inverse of this problem arises when we get incompatible standards. One of the classic examples of this is the old VHS/Betamax competition, in which the superior product (Betamax) lost out. While it’s easy to shrug your shoulders and say, well, that’s the market for you, we must remember that this competition required consumers to make uninformed guesses as to which technology would prevail; when the matter was decided, consumers lost millions of dollars. This was not a shining moment for the market economy. Capitalism sometimes screws up.

We’re now going through exactly the same thing with the HD/Blu-Ray competition. Once again consumers are being asked to flip a coin, take a gamble, and see if they win or lose. Some will win; some will lose. And there’s no way right now that any consumer can make a reliable judgement as to which technology to pick.

If we had the same market-based approach in telecommunications, we’d have the same problems, only larger. For example, consider cellphone service. You cannot readily mix and match hardware with service providers. If you buy handset X, you’re probably stuck with service provider Y. Apple’s iPhone is a wonderful technology, but you can’t use any service provider you want. Moreover, once you’ve chosen a service provider, your service area is limited. You can’t be certain that service provider Y will provide good service in the same areas that service provider Z does. The ironic result is that a free market for service providers creates a less-than-free market for consumers.

OK, time to stop and let you digest this.

Of course, this still raises really wild problems. What if you buy up several deeds, combine them, use them as a whole, and then later break them down and sell off pieces? Would this create legal nightmares?

I don’t see why. You would essentially be squatting in that intermediate zone, so it’s not like any rights to that area would get transferred during the break up. It would simply revert back to the original arrangement, because then the new owner would be injuring someone by encroaching on the intermediate zone.

No, because you can’t have contiguous areas with the same channel, because of problems with S/N ratios. Go back to this statement in my previous comment:

“But consider this: nobody can have any rights to channels 2,4,5,7,8,10,11, or 13 in the south central region. We’re throwing away 2/3 of the spectrum in order to obtain clearly defined property rights.”

And read the preceding paragraph to see why.

I just realized, though, that you could get away with it by exceeding the terms of the deed but insuring that nobody has standing to sue you because you already own the other rights. In other words, even though nobody owns the rights to your channel in the intermediate zone, if you own the rights on the two peripheral zones and transmit into the intermediate zone, nobody else would be injured by your action, and so you could technically violate the deed without any problems.

Of course, this still raises really wild problems. What if you buy up several deeds, combine them, use them as a whole, and then later break them down and sell off pieces? Would this create legal nightmares? I need some time to think this through.

Of course, the FCC’s registry would be a lot more complicated, but it’s theoretically possible.

I’ll skip this for now because you acknowledge that it’s possible. Feel free to come back to it if we overcome the as of yet unstated deal-breakers.

But here’s the rub: you can’t combine deeds to assemble something bigger. If you buy up the rights to Channel 2 in both the south bay and the mid-north bay, that doesn’t give you any rights in the central bay or mid-south bay. There are no economies of scale.

And I can’t put up houses, let alone hotels, until I acquire Tennessee Ave, New York Ave, and St. James Place.

If you acquired the rights to Channel 2 in a properly shaped contiguous area, you would have an economy of scale, right?

The owner of this deed has to right to transmit in the region marked on the map in the frequency band from X MHz to Y MHz, to a maximum power flux of Z W/m2. The transmission may not exceed a power flux of A W/m2 in any other frequency, nor may it exceed a power flux of B W/m2 outside of the marked region.

Now, this is a simple, clean system that is easily measured and enforced. Unfortunately, it raises practical problems. The power flux received at a point distant from an antenna depends not only on the power level and the antenna design, but also on many factors outside the control of the transmitter: humidity, air temperature, the state of ionization of the atmosphere, and the solar wind are the biggest factors, but there are more. Therefore, in order to remain within the legal bounds defined in the deed, the transmitter will need to operate at a power level less than optimal.

But it gets worse. At the boundary between two regions, if both transmitters are operating optimally, then the signal strength of the two transmitters will be the same — meaning that a person on the boundary can’t get a good signal. Depending upon the type of transmission we’re talking about, we’ll need a S/N ratio of at least 4:1 to get good reception. But in order to accomplish that, we need the dead zone between two transmitters to be equal to half the separation between them — which means that we can only cover about one quarter of the total area! Three quarters of the people in the area don’t get any reception!

The way around this is to overlap different regions with different channels. That’s why there are 13 VHF television channels even though for most of the time there were only four main networks: ABC, NBC, CBS, and PBS.

Here’s a simple illustration of the problem. Let’s talk about the San Francisco Bay Area and treat it as if it were one long, continuous target area. We’d divide it into, say, five zones: North, mid-North, Central, mid-South, and South. For the South Bay, we’d set up ABC on channel 2, CBS on channel 5, NBC on channel 8, and PBS on channel 11. For south-central, we put ABC on channel 3, CBS on channel 6, NBC on channel 9, and PBS on channel 12. For central, we use ABC on 4, CBS on 7, NBC on 10, and PBS on 13. For mid-north, we start the cycle over, with ABC on 2, CBS on 5, NBC on 8, and PBS on 11. Now, the mid-north transmitters might interfere with the south transmitters, but they’re so far apart that the interference can be handled.

But consider this: nobody can have any rights to channels 2,4,5,7,8,10,11, or 13 in the south central region. We’re throwing away 2/3 of the spectrum in order to obtain clearly defined property rights.

Now, one solution to this problem is to use directional antennae. These allow you to crowd transmitters closer together without getting much interference. But these raise a whole host of new problems. You see, the ability to control the power flux at any given point in a region depends upon the wavelength. Longer wavelengths mean less control, and more interference problems. So now you have to rewrite the deed to take into account the differences in wavelength; deeds for lower frequencies need laxer standards than deeds for higher frequencies.

Then you have to take into account topography. Are there any mountains on which to site antennae? What areas can be reached by antennae on those mountaintops? What about intervening mountains that might block transmissions? The Hollywood hills in the LA basin block higher frequencies, but not lower frequencies. So now you need to write a custom deed for each and every transmission region, marking out its limits on the map, and defining the power density standards. And you have to do a separate version for each and every frequency band.

Now, this isn’t impossible, it just takes a big bureaucracy to nail down the specifications, and this is what the FCC spends a lot of time doing. We could have made the FCC not a regulatory agency but a registration agency, like the land registration offices in every county in the country. They keep records of who owns what. Of course, the FCC’s registry would be a lot more complicated, but it’s theoretically possible.

But now let’s imagine implementing this system and letting it evolve according to market forces. Every year there are hundreds of transactions in which people transfer ownership of narrowly defined chunks of spectrum and territory. Now, you can always break down a deed by frequency — breaking your bandwidth into two chunks and selling the two chunks as two new deeds. But here’s the rub: you can’t combine deeds to assemble something bigger. If you buy up the rights to Channel 2 in both the south bay and the mid-north bay, that doesn’t give you any rights in the central bay or mid-south bay. There are no economies of scale. It doesn’t buy you anything special.

Imagine land use laws that permitted landowners to break down their parcels into smaller and smaller parcels, but never permitted them to combine smaller parcels into a bigger parcel. What would you end up with in the end? An impossible situation!

Gee, and I still haven’t gotten to problems of capital intensity, standards creation, and “winner takes all” in technology. If you guys still aren’t convinced, I’ll get to them.