For fun This Onion article is hilarious. In what CEO Bill Gates called "an unfortunate but necessary step to protect our intellectual property from theft and exploitation by competitors," the Microsoft Corporation patented the numbers one and zero Monday.

HTML Purist Mark Pilgrim writes this well considered piece on what HTML is actually for, and why it's so successful. My favorite quote: "whole idea of replacing desktop applications with web-based applications has set the entire field of user interface design back at least 10 years". This issue will become increasingly important as application features get distributed over an (internal) network.

Auction delay To date, spectrum licensing has been the single most damaging government intrusion into the technology world. While Hollings (Dem, South Carolina) tries to top this with his outdo his Disney bought CBDTPA, the spectrum crowd still has a few tricks up its sleeve. There are moves afoot to force the FCC to actually hold the spectrum auction that's been delayed for two and a half years. The CTIA wants to postpone this indefinately until they work out a "more logical" spectrum plan.

The thing is that there is nothing more logical than a spectrum auction. Those most able to profitably use the spectrum will bid the highest for it, and if they get it wrong they can then sell those rights to other people. The market will decide which uses it most values, and those holding spectrum inefficiently will bear the opportunity cost and so sell the spectrum to those who can use it better. Spectrum is a rival good, so strong property rights and a free market will allocate it best.

Protectionism good Winterspeak started off as a free-trade and economics blog, before focusing on law, economics, and technology. These worlds collided today as Congressional disaster and wholly-owned subsidiary of Disney, Inc, Ernest F. Hollings posts a protectionist screed in the NY Times. Better known for trying to cripple all technological innovation, Hollings' today trots out the usual hackneyed protectionist argument saying that protecting workers helps the country (which ignores, as usual, the larger social loss on consumers brought on by higher prices). "Economics in One Lesson" is a well written book that explains why this thinking is dead wrong (it's 50th Anniversary Edition just came out, which also shows how old Hollings' ignorance is).

I most enjoyed Hollings' argument against Fast Track authority, that it transfers "power over trade to the executive branch and favored corporate interests." But as someone trying to ban the software industry at the behest of his Hollywood masters, at least he's talking about something he knows here.

Complementarity cuts both ways I was reading through Murphy's testimony and came across this gem:

In economics, complementarity always runs in both directions. Lower price for gas, more demand for cars. More cars, more demand for gas. It goes bothways.

This is just the flip side of the argument made by the court that the emergence of middleware generates demand for more operating systems. You can't have one without the other. The two reflect the same underlying economic principles. But people don't always realize this one even though they realize the first one.

I thought there had been a sudden explosion in operating systems recently, and now I know why--the growth of middleware creates demand for specialized operating systems, so they appear. I'm betting that this is what Tim O'Reilly's emergent Internet operating system will look like, an archipeligo of middleware linked by many languages and protocols. Nice.

Spoke with Murphy I spoke with Kevin Murphy this morning about his Microsoft testimony. His position is that Microsoft's monopoly position is so strong, neither Java nor Netscape posed a real threat to it. Therefore, Microsoft did not cause their demise because they were doomed anyway.

Sadly he's on firm ground here. Microsoft owns the desktop, and mandatory unbundling of middleware would freeze the Windows OS. This doesn't increase welfare.

Murphy also said that forward looking regulations end up creating as many problems as they solve, and this has been true in telecom and utility regulations. Since the operating systems in the future will hopefully be very different from what we use today, it's hard to think of rules that would stay relevant.

The consensus at Chicago is that Kottler-Kelly will OK the proposed settlement, with hopefully fewer loopholes. I also expect more anti-trust prosecution against Microsoft in the future.

Code enforces cartel contracts Cartels are good for an industry and bad for consumers because they keep prices high. But cartels are also unstable because every company has an incentive to cheat and try to grab marketshare by lowering price. It's tough to punish cheaters because throwing them out of the cartel means they keep on cheating, and inviting them back into the cartel means they can just cheat again. Punishing the cheater by starting a price war just hurts everyone (except consumers, who do quite well).

You can't draft private contracts enforcing cartel rules because such contracts are illegal. However, suppose you built such a "contract" into the code of a system--there would be no paper-trail, the cartel would be enforced, and you would remove all incentive to cheat.

CSS region encoding on DVDs does exactly that. Content owners want to price discriminate between different countries, but how do they keep distributors from reimporting DVDs from low-priced to high-priced countries? The MPAA did this by building regional encoding into the technology itself, DVDs imported from Europe can't play on US DVD players, and vica versa. DVD player manufacturers aren't too happy about this as it means 1) their costs go up and 2) their markets shrink, so to placate them they get a royalty from DVDs sold by region (giving them a cash incentive to support regional encoding) AND their machines can switch between regions 6 times (or something like that) meaning people can move across the Atlantic a few times and their equipment still works. Remember, the point of regional encoding is to stop re-export in DVDs.

All of this is collusive price-setting and would be illegal if written in paper form. But build it into the code, and you get a stable cartel. Nice.

War of analogies The motion picture industry successfully censored 2600 for posting DeCSS code under the DMCA in 2600 v Universal Studios. The case is now on appeal.

I heard Chicago law prof. Geoffrey Stone talk about the case and its First Ammendment ramifications today. Basically, he said that if a law incidently blocks speech, but is fundamentally not about speech at all, then the Supreme Court lets the law stand. This is sensible since almost any law can, in some situations, block speech, and the Supreme Court doesn't want to be buried under millions of First Ammendment challenges. The second point is that if "words" are essentially like keys to a lock, they can have their functionality prohibited because prohibiting functionality is incidental to banning content.

DeCSS, then, becomes a case of two duelling metaphors. Is it like a key, passed on from one accomplice to another to further an illegal end? Or is it like a book (or movie) describing a homicide in painstaking detail, which is valuable both as entertainment but also teaches someone how to commit crime? The former does not enjoy First Ammendment protection, but the latter does.

I think a good way to attack this case is to argue that CSS is not about preventing copying at all (which it isn't), it's about enforcing industry collusion to geographically carve up sales territory. This doesn't declare the DMCA unconstitutional, but it does carve out an exception for DeCSS and other code-backed cartels.

Digital TV remains dead Digital television was cooked up by Zenith to prop up the US television manufacturing industry (i.e. itself) in 1989. The FCC allocated spectrum it didn't have to this new medium "at the behest" of Congress for free, never mind if people actually wanted to use spectrum for this. The shill used to sell this was "higher resolution" although since the soap and cornflake companies that actually pay for broadcast TV don't care about resolution, all we'll get is "more channels". If we get anything at all.

To date, digital TV has been an utter failure. No one wants to buy the expensive, useless digital sets, so no one wants to broadcast the expensive, useless digital signals. Goaded by Hollings' (Dem S. Carolina and wholly owned subsidiary of Disney, Inc) nightmarish SSSCA bill, electronics manufacturers have agreed to build digital restrictions management systems into the sets. This prevents recording TV signals. If digital TV was less utterly pointless, I'd care.

Murphy and Microsoft Kevin Murphy is a Chicago economist's Chicago Economist. He testified for Microsoft yesterday, arguing there was no causal link between Microsoft's actions and Netscape (and desktop Java's) demise. The DoJ argued the feeble settlement was neccessary because they didn't show causality. I hope to speak with Kevin about this next week, and I'll tell you what I find out.

CD Price fixing Will Cox pointed out a price fixing debate on CD sales.

What the charts actually show is that the elasticity of demand for CDs is about -1.5, which seems reasonable. The price fixing charge is better supported by all the labels having the same suggested retail price. Remember, the RIAA, operating as a cartel, offers CDs to (commodity) retailers at about $10 each. Anti-"red lining" laws enforce the cartel's agreement not to cheat in their collusive practices, so none of them give retailers price breaks. Retailers sell at marginal cost to consumers.

Eldred v Ashcroft Morton David Goldberg works for Cowan Liebowitz and Latman, a NY law firm that represents copyright holders. He was originally going to represent the IP section of the American Bar Association in arguing for the Sonny Bono Copyright Extension act, but now it seems he's just representing himself. In this interview with Salon, he argues both that copyright extension is good, and that this is something Congress should decide, not the Supreme Court.

He's dead wrong about the former: its hard to motivate the deceased to create new works, it's hard to have created works be more created, and it's hard to claim retroactive extension is "prospective" and keep a straight face. But he's kind of right that Congress should be held accountable for bad laws--they passed them and they should change them. But Sonny Bono is clearly a case of regulatory capture and that's why we have a Constitution, and a Supreme Court.

I've been encouraged by the (temporary) death of Holling's SSSCA, the mainstream press articles on the social harm caused by current copyright law, and the Supremes agreeing to hear this case. Congress has a long history of passing, and then overturning, bad legislation.

Used books Bezos has an open letter defending peoples right to buy and sell used books. Selling a book after buying it lowers its real cost to the buyer. The correct response for publishers, if this effect becomes large, is to raise the price of new books. Publishers will capture their old royalty up front, and sell fewer books. The aftermarket price of books will rise as the quantity of books available (through first sale) goes down. So, publishers need to capture more of the value upfront, as efficient aftermarkets gut post-sale monopoly markups

The extreme case is the future of music: recordings from live shows freely distributed online.

Moral rights This article explains what moral rights are, but not why they are bad. In Europe, the original artist retains some control over any work he produces (although I don't know if "work for hire" clauses work there, in which case the point is moot). This can only limit the value of the art to a buyer, making him less willing to pay for the art and so impoverishing the artist. And it ignores the economics behind copyright, which creates dead weight loss by artificially keeping price above marginal cost.

Hailstorm is toast Microsoft couldn't get any company to sign up for Hailstorm so has shelved the project. They're now thinking of offering it as an identity platform companies can use instead of developing their own. This is a massive step-down from their original plan which was to establish control at the user level.

Cripple the aftermarket Traditionally, book publishers price discriminate by releasing expensive hardbacks first, cheap paperbacks later, and inconvenient library access. Marginal cost is constant across all three segments. But Amazon (and eBay/Half.com) takes friction out of the aftermarket by making it easier to buy and sell second hand books, which reduces dead weight loss and transfers surplus from publishers to consumers. Publishers, of course, want this banned.

Carping about CARP The recording industry cartel set web-radio rates at 0.07 cents per song per listener for over the air broadcasters, and 0.14 cents per song per listener for pure internet broadcasters. This has upset people.

I won't quibble over the price, but predictably web-radio says they're too high, and the RIAA says they're too low. The real meat is in the rate *structure*:

1) Per listener
Broadcast radio does not have to pay anything per listener. Broadcasting over the air is a public good, in that marginal cost is zero, and audience-tracking is hard, so not having per-listener fees made sense. But web-radio is *not* a public good as bandwidth requirements rise quickly with listeners. Add this to easy audience-tracking and a per listener fee does not seem so unreasonable--it increases marginal cost but does not distort the cost structure.

2) Standard
The RIAA uses regular radio broadcasters as an advertising channel and pays *them* to play certain songs. This sleazy industry practise is called "payola." Web-radio argues that since broadcasters don't need to pay they shouldn't either, but sadly this argument doesn't hold. Radio stations pay what they're paid to play, while web-casters (proudly) play whatever they want. Eclectic variety doesn't help the RIAA generate maximum rents (that they can then appropriate) from bands they own, so this does not help them make more money. Cash may flow from RIAA to radio, but it'll flow from web-radio to RIAA if web-radio wants to make its own programming choices.

So how much cash should flow? Presumably the big labels want to charge a high price, while little labels (who have trouble paying for radio placement) may be willing to charge less. Maybe a lot less. But since the rate is *standard*, you won't get price competition between labels, so the cartel gets to keep its control over distribution. But suppose some bands were so desperate to be broadcast they'd be willing to forgo royalty payments altogether?

3) Mandatory

Sorry, you can't do that, CARP royalties are mandatory (I couldn't download the report, so please tell me if I have this wrong).

And now it's clear what CARP is really all about.

CARP isn't about rates at all. It doesn't really matter what the rates are set at so long as they're *standard* and *mandatory*. If you think about what the RIAA really owns that's scarce, it isn't singing talent (which is common and fungible), but rather the complementary distribution channels and promotion apparatus. Napster gutted their distribution monopoly, and web-radio guts their promotion monopoly, which is the *real* reason they're blocking both.
Link to this column

Microsoft will remain insecure NYTimes has a piece on how Microsoft is becoming more "secure". But security isn't just refactoring sloppy code, it's keeping systems simple and focusing on secure processes. Complexity is anathema to security, but Microsoft routinely adds complexity to products to make them harder to immitate (and, contrary to their assertions, harder to use).

Another reason to love Maxygen There's a very old and cited legal ruling (Baker v Seldon, 1879) that stipulates how copyright law protects only the expression of ideas, not the ideas themselves. A while ago I wrote how Maxygen wanted to copyright DNA by converting it into music. Their argument is that the underlying genetic sequence, once expressed as music, falls under copyright and cannot be legally accesed through circumvention, as per the DMCA.

Now, converting a DNA sequence to music, or using *any* encryption technology, is essentially expressing an underlying idea in a particular way. But by making circumvention illegal under the DMCA, copyright extension is de facto extended to the underlying idea, in violation of Baker v Seldon. This case, which has held up for over 120 years, went so far as to say that when expression and the underlying idea were unseperable, copyright was not applicable at all. If Maxygen goes ahead with this (possible), and is challenged in court (likely), it will seriously challenge copyright law, or the DMCA, or both. Nice.

Valenti Jack "Freedom is the enemy" Valenti clearly outlines Hollywood's goals in its ongoing jihad against the personal computer. He wants:

1) Policeware that checks for broadcast flags and keeps digital TV content off the Internet.
2) Banning all bitstreams that have no flag at all (to block people from redigitizing analog content. He refers to this as "plugging the analog hole" which I find delightfully colorful.)
3) Prohibiting individuals to self publish ("the persistent and devilish problem of peer-to-peer.")

The policeware system is designed to cripple the aftermarket for digital content. After all, advertisers put stuff on TV you want to see so they can show you stuff you don't want to see. So why should they care if their ads are distributed through more channels? Advertisers don't, but Hollywood does because it also sells its content directly to end-consumers via Blockbusters and unless they cripple the digital aftermarket, they will not be able to price discriminate between these two customer sets. (This is kinda how regional encoding in DVDs work)

Banning all unflagged bitstreams means consumers will be barred from producing their own content. A digital camcorder doesn't know if it's pointed at your nephew, or to a cinema screen. A microphone is indifferent towards a garage band and Celine Dion. Holywood's answer to this is just to assume all bitstreams are "stolen" and ban them all. This is insane. But no more insane that making it illegal for computers to speak with each other, which is essentially what the last point entails.

You'll note that text based content providers aren't moaning about rampant copying. They're free to lock up their content as much as they like, but what they do is irrelevant since readers have backward integrated into the content creation field via weblogs and are happy giving the stuff away. (Musical successes will do the same thing for recording). But Hollywood is on crack to think their business is threatened. Video transmits at 180Mbps uncompressed, and let's say 90Mbps compressed (with loss). At 0.3Mbps (which is what the FCC defines as "broadband") it'll take 750 hours to download a 2.5 hour movie. Even if you improved everything by three orders of magnitude it's still faster to go to Blockbusters.

Yet more media consolidation Peanut Gallery brings up the "diverse viewpoints" again claiming that the anti-"piracy" screed in this year's Grammys (which I did not see) and subsequent lack of alternative reporting on CNN demonstrates how consolidated media is systematically attempting to manipulate public opinion. I would say it demonstrates how little people care about abstract technical debates so long as they can still download mp3s. Oh, and newspapers were full of stories on this stuff. Don't assign to brainwashing what you can assign to apathy.

He also brought up Time Warner distribution blocking ABC content as an example for why my economic rationals did not hold. But I never said content owners and distributors wouldn't haggle over price, I just pointed out that even a monopoly distributor would let other people use its pipes for the monopoly price. Businesses aren't reliably smart, but they are reliably greedy. Thanks to Stanton for some good email coversations. And now I'm putting this dead horse to bed.

Media consolidation revisited Stanton responds to my piece on why owning content and pipes does not matter. He argues:

IMO, the costs to society of a few companies controlling the message far outweigh any economic benefits of consolidation...[And] how does a company make as much money as possible? Simple, by gaining sole control of the market, by becoming a monopoly.

I think Stanton raises some commonly held arguments, so lets take them individually:

1) "Having only a few companies control the message is bad for society."
I can only think of two sorts of "control" an integrated media company has, 1) refusing to distribute someone else's content or 2) favoring their own content. As I explained earlier, if you factor in opportunity cost (which you should) an integrated company has no incentive to do either. So if the integrated company was maniacally evil they might be willing to lose money by "controlling" the message, but so long as they're just maniacally greedy, we're OK. And I think we're pretty safe on that score.

Moreover, the number of information outlets has been exploding for the past thirty years, and especially with the Intenet, there are more sources of "message" now than there have ever been before. I know from my law classes that practical reality is not popular in legal circles, but worrying about narrow content choice in this day and age seems pretty surreal. (Finally, there are also cases where splitting an audience too narrowly actually reduces media options by making that segment unprofitable to serve. But let's not get into that.)

2) "A company makes as much money as possible by being a monopoly."
I agree, but the details here matter a lot. The value of a monopoly is determined by the elasticity of demand for the monopoly product, which in turn depends on the cross market price elasticity between that market and close substitutes. There are many close substitutes for broadcast TV, such as cable and satellite. There are even more further substitutes, including radio, newspapers, and the Internet. While the three big network broadcasters (ABC, CBS, NBC) continue to collusively set ad rates, their fragmenting audience share makes that harder and harder. (Note: if companies compete on price, you only need two players to reach perfect competition. That's why collusion is so important in such industries).

And you have to remember, broadcast networks' customers are companies that advertise with them, not consumers who watch their shows. Broadcast networks *real* use of monopoly power is limiting *national* content distribution channels to keep their *national* audiences concentrated and *national* ad rates high. And cable, along with satellite distribution, has been gutting this for years.

Which brings us to the central insight--the three broadcast networks collusively protect their monopoly by limiting entry into the *national* TV advertising market. But the "anti-monopoly" concerns that oppose mergers like Hughes and Echostar ironically end up *protecting* the national TV advertising monopoly in the name of "not wanting media to get too powerful."

The broadcast cartel understands this and disguises their opposition to entry under "we must protect media diversity" rhetoric. This stuff's really obvious if you know how to look for it. For example, EchoStar is arguing it should be allowed to broadcast local TV across the US. What Hughes/EchoStar *really* want is a national broadcast network to compete with ABC, CBS, NBC. Local news is as unimportant to them as it is to the huge majority of viewers who don't watch it (ratings don't lie). So why is Hughes spinning this as a "we want to broadcast local news everywhere?" Because the broadcast cartel has convinced people local news protects media diversity and uses them (and that) to block anything that undermines *actual* competition in the *real* market.
Link to this column

Cheap CDs in Mexico There's a NYTimes piece on cheap CD copies in Mexico. One shopkeeper argues that Mexicans buy copies because they can't afford expensive original CDs--so why doesn't the recording cartel lower prices across the border? Because these cheap CDs would be reimported back into the US undercutting the prices there (if the recording companies could price discriminate, they would). Regional encoding is how the movie and console cartels price discriminate between countries, but there is no regional encoding on CDs (yet).

Carly saves her job Looks like Compaq/HP has gone through and Hewlett Jr. is out. Carly gets to keep her job while the merged company flounders, and then it will fall apart, and she'll step aside to spend more time with her family. Most mergers fail, this will be no exception.

Cripple the aftermarket A while ago the recording industry got Garth Brooks to try and ban second hand CD sales. This failed. Note that banning resale would not effect new CD purchases--if I can sell my CDs second hand then the cost of a new CD falls to its sale price minus it's resale price, so I can afford more CDs overall and the revenue remains unchanged. But eliminating the aftermarket by banning resale does allow recording companies to price discriminate by charging different amounts to different people (right now they simply collude to charge the monopoly price).

The recording industry wants to price discriminate when it sells digital music, but here the aftermarket problem is even more acute (as people are happy redistributing at marginal cost). I'd say 80% of current RIAA funded legislation is focused on crippling the aftermarket for music sales so they can price discriminate, 10% is on protecting their cartel, and the remaining 10% is on appropriating rents from artists. (link via Government Monkey)