Friday, June 28, 2002

Why are Mac's piling up? The iMac, once in short supply, is now gathering dust on shelves. Why? Part of the reason is the price increase -- if things become more expensive, people by less of them. And if customers anticipate a price, they will delay their purchase until the price comes down. Finally, I think people are waiting for the release of OS X.2 or big annoucements at MacWorld NYC, so again are delaying their purchases. Coase, who is utterly brilliant, wrote about the effect of strategic buying decisions on pricing power, and that's working against Apple here. Moreover, the internet may be making this worse as the Mac faithful know much more about what to expect than they did in years past.

Quick links Two good articles from the Register today, the first discusses an asanine report from the BSA on al-Qaeda's "e-terrorist" plans and the second looks at Palladium, Microsoft's (patented) DRM OS.

Thursday, June 27, 2002

Lotus Bloat More struggles with my Lotus Notes client today. While you can use Notes to create all kinds of sophisticated applications, it handles basic email poorly. We were all complaining about the sluggish client, but it turns out this was because we were working off the networked database, not a local replica. Why the system was configured this way, I don't know. So we make a local replica and things are better, until I go to a meeting with my laptop. It turns out creating a local copy of email does not also make a local copy of your calender or to-do list. I can imagine some fevered engineer now: "what about the class of user who wants to have their mail everywhere, but only wants to see their calender when they're networked?" POP3 does a better job of meeting customer needs.

Wednesday, June 26, 2002

Console market You have to wonder what Microsoft is thinking about it's X-Box. Until Microsoft's desktop monopoly helps it enter the console market, it's as if someone lent me $5B to compete with against the PS2. And despite what dumb business journalists think, the Borg's cash hoard does not give it any special advantage--as a MSFT shareholder, I want to see the Beast give its money to me through share buybacks or profitable businesses, not squander it competing against Sony and Nintendo. Just because Microsoft has money doesn't mean it can waste it.

Tuesday, June 25, 2002

Is Palm OS toast? While some think Be's NeXTish reverse take-over of Palm OS means the company's been saved, I'm less convinced. Be failed because it was run by insane engineers who seriously thought OSes didn't need applications to ship and what users *really* wanted was to play 3 DVDs AND move their windows around simultaneously. Why this flavor of insanity should help Palm, whose major flaw was to ignore the corporate market, I don't know.

Palm was the hardware company whose user-centric philosophy came closest to Creative Good's. It succeeded where an entire industry (Newton etc.) failed and owns about 75% of the handheld market. But in corporations, end customers don't make the buying decisions and IT is out of touch with basic user needs, so they push the bloated and expensive PocketPC on their long suffering users, giving MSFT all of the remaining 25%. Palm OS 6 (or Be 2.0 Micro edition) will give handhelds plenty of oomph, but they'll be running on expensive, greedy ARM chips (not trailing edge cheap-as-dirt Dragonballs) and probably ship with no apps. How this will help Palm, I cannot imagine.

Yahoo! RIAA deal was designed to stifle competition Mark Cuban, who sold to Yahoo!, has stated in writing that the deal Yahoo! signed with the RIAA was designed to exclude small webcasters from the market. Such a contract would not be feasible if the RIAA was not a cartel, so let's hope for toothy anti-trust investigation soon. It also puts the Library of Congress in the embarrassing position of giving an anti-competitive royalty regimen a government mandate in CARP, and if people raise enough of a stink I can see that decision being repealed.

As it stands, the RIAA has the worst of both worlds: they aren't charging big webcasters enough and have embarrassed the government by tricking them into shutting small webcasters down. I'll be interested to see where this story goes from here.

Monday, June 24, 2002

Getting it exactly wrong I'd like to say that nonsense like this ridiculous spiel that draws parallels between radio and the Internet are rare, but they're common as muck. The article calls foul because Congress declared radio airwaves to be used for "public interest, necessity and convenience" and then dished them out to commercial entities instead of nonprofits.

Typically, the author does not realize that the "public interest, necessity and convenience" regime itself guaranteed public harm by having government decide how spectrum was to be used, instead of markets. Radio spectrum should have been auctioned off to the highest bidder, ensuring the public is justly compensated for private use of its property, and that the resource would go to whomever could use it best. If nonprofits wanted to cough up the dough, they'd be free to do so just like everyone else.

The entire notion of "public interest, necessity and convenience" came from New State Ice Co. v. Liebmann 285 U.S. 262 (1932) where an incumbent ice company successfully blocked an entrant claiming that any resulting competition would be "ruinous." My professor, Randal Picker, felt there was some validity to this claim, but it's bogus. (via Archipeligo)

AOL should stay a dinosaur A reader wrote in commenting on my AOL post saying
"the only way for AOL to convert from being the worlds largest dinosaur ISP would be for them to figure out how to offer a conversion to Broadband to all of their current subscribers, and figure out the business model to pay for it second."
I think this is a common position, both within the tech industry and without, but it's not one I agree with. AOL is a public company, and its job is to make money for shareholders, not inject cool new technology into the market. It's widely supposed that tech companies make money by injecting cool new technology into the market, but the fate of Apple, Be, NeXT, and the enormous number of cool tech companies that struggle go to show that's just not true. AOL should focus on making money and leave public works to the government.

Saturday, June 22, 2002

Broadband Wireless There was a time I used to roll my eyes whenever i heard the phrase "broadband wireless" (it sounded to me like "visualize synergy") but this Economist article does a good job of covering emerging wireless technology.

Friday, June 21, 2002

Complaints about CARP It's hard to fault journalists for not having a clue when none of the "expert" commentators do either. The US Government has set a rate of $0.0007 per song per listener mandatory royalty rate for web-casting, and small web-radio stations are complaining it's too high. The truth of the matter is that only a market can determine the correct price of something, and the real problem with the Government's rate is that it's mandatory. Some artists might want to set a lower rate for their material (say, $0) or even pay web radio to broadcast it, but with the current structure there is no way they can compete. While the Government claims to be mimicing "market rates", what they've actually done is destroy the market by eliminating competition through mandatory rates. Instead of squabbling over what the price should be, web-radio should have insisted the final rate would be a ceiling and let the market clear.

Wednesday, June 19, 2002

AOL It's late, so I'll keep it short. Had a good chat with a couple of people about AOL. Turns out that their landline business is pretty profitable, but they lose money on broadband customers. Wallstreet is pushing them to convert more people to broadband though, even though this hurts earnings. Analysts are nuts.

Tuesday, June 18, 2002

Who radio conglomerates really threaten Here's a typically misguided piece on how recording companies (NYTimes) are teaming up to stop radio consolidation. The RIAA prefers small local stations because 1) they're cheap to bribe and 2) they're in weak negotiating positions. Regional networks like ClearChannel thwart this because 1) they charge higher rates and 2) they're in more powerful negotiating positions.

A *seemingly* peculiar twist to this legislation is the labels' call Congressional limits on the amount labels pay radio stations. The reason they're asking for governmental intervention is because such a (private) contract is collusive, and therefore illegal under US law. But the recording industry has a long and rich history of government granted monopolies, protection, and injunctive relief from the future and there's no reason that should change now. Amusingly, the same folks who decry the RIAAs outrageous legislative actions in the technology realm (DMCA, AHRA, SSSCA) will probably support this piece of monopolistic, pro-RIAA market-rigging, thinking they're somehow standing up for the little guy when really they're just supporting very incumbents they claim to detest.

Monday, June 17, 2002

Joel on Micro Microeconomics is a beautiful thing, and I go to Chicago, so I really get to appreciate it in all its glory. I'm glad Joel has picked up some micro and applied it to the software biz, but let's take his thinking a little further.

Two goods are "complementary" if demand for them rises and falls in tandem. For example, if the demand for right shoes goes up, demand for left shoes will inevitably go up as well. Joel is right in saying OS and hardware are complements, so if the cost of one falls through the floor (boosting demand), the demand for the other will rise as well even though it's still at the old (high) price. Therefore, all companies want to commoditize away the complementary good and become "left shoe monopolists".

Or do they? Economics is about trade-offs, and commoditizing away complements has the bad effect of removes their incentive to invest in boosting demand. They may slide down the demand curve, but they lose the ability to shift the demand curve out. Integrated complements can jointly coordinate and invest in activities that increase demand for the product as a whole. Coca-Cola, which makes concentrate, gives its bottlers local monopolies so they properly invest in downstream retail relationships. Content companies invest in cable companies to promote local advertising for premium programs. McDonalds operates transient franchises on motor ways to make sure they're well run. DeBeers doesn't believe jewelry retail can do much to increase demand for diamonds, so forms appraisal agencies to commoditize them away. Apple coordinates hardware and software to introduce innovations like USB. Microsoft insists OS and middleware/applications must be integrated so it can offer the best computing experience. The list goes on.

Firms need to figure out whether they're best served by commodity complements (that move down the demand curve) or integrated complements (that can shift the demand curve out) and then do that. The entire computer industry except Microsoft wants desperately to commoditize operating systems so it can get back to innovating, and GNU/Linux is their best bet.

Sunday, June 16, 2002

Quick update Archipeligo has a nice response to my recent post on the strangely weak cross price elasticity between CDs and mp3s. And to think I never knew Daniel was so musical.

Friday, June 14, 2002

Quick links Here's your typical alarmist article about broadband consolidation and how this will cripple media freedom, blah blah blah. I've gone on at length about why all this is misguided, so will just point out the (falsely) apocalyptic language here. The piece is full of stuff like "broadband access providers can now deny other DSL providers access" without pausing to think why they would deny access. Even if the pipe was a monopoly, it would simply charge the monopoly price, not deny access altogether, and right now pipes are hardly monopolies. People worrying about commercial censorship tend to miss the obvious fact that the only entity in this business is their beloved government.

Finally read Jane Galt talk about Kevin Murphy. It's an honor to be at a place like Chicago where the utterly brilliant Gary Becker may actually not be the biggest genius in the house, and the biggest genius may actually have to stand inline behind someone else to get a Nobel Prize. To me, being in Chicago is like being in some kind of heaven.

Thursday, June 13, 2002

Almost right Here's a reasonable article that says since file sharing is so prevalent, but album sales are pretty much where they were, downloaded mp3s must be poor substitutes for CDs. This is a remarkably sensible observation, but leaves you wondering why.

I don't know the answer, but I'm guessing that demand for music is very elastic while demand for CDs is less so. This means I'm willing to listen to a song (for free) if someone suggests it, but I'm not going to buy a $15 CD just to check something out. But why would CD buyers continue to buy CDs when they can download everything for free online? Part of the answer is that CDs are more portable than mp3 players (iPod excepted, other portable mp3 players suck) and so are important for music on the go. A CD where every song is worth having is probably better value than the time spent finding and downloading them online. Or maybe people just like having CDs.

IBM rocks Had a great day at Lotus today, chatting with Gary Cohen and Sean Martin of Sash fame about IBM and where it was going. Some of it was around distribution, which sounds kind of boring but often turns out to be the scarcest complement, and so ends up making all of the money. IBM has distribution in spades, but this also makes it hard for them to capitalize on their (remarkable) research -- their channels are specialized and ill suited to new, unproven products. This rigidity is why start-ups have any hope at all.

Wednesday, June 12, 2002

Pricing Bandwidth Broadband providers are thinking about capping or metering customers, supposedly to stop freeloaders. But this is wrong, so let's go through the economics:

Broadband provision is a high fixed cost, low marginal cost business, so while average costs are high, marginal costs are low. Also, while service provision is cheap, each customer is expensive to get hooked-up and online, so setup charges eat into lifetime value, and hence, profitability.

The article states that "all the cash cable companies spent sinking broadband, only to see slow uptake, means they have to raise price now" is wrong in every respect. Deployed cable is a sunk cost so shouldn't figure into any pricing decision. While business often make the mistake of using fully loaded costs in pricing decisions, pricing below average cost is a signal to exit the business, not raise prices, because higher prices will just hurt profitability further. Secondly, broadband uptake is a healthy 12% per quarter, or 57% a year, which may be slow compared to fevered executive dreams in 1999 but is still pretty respectable. Given how order congestion increases the cost of adding new users to the system, and how low broadband penetration currently is, higher flat-rate prices are probably meant to slow adoption, not harvest the current installed base. When you have a small customer pool, your profit depends on adoption rate because 1) you have many more potential customers than actual customers and 2) new customers become regular customers.

The article also misses the its own internal contradition, namely that if unused capacity is why companies have to raise prices, why should they mind if a few users tie up extra bandwidth? High-use customers aren't a problem, they're just an opportunity to charge higher prices to a segment that will probably pay them. And this brings us to the real story.

One way to tackle high fixed cost, low marginal cost businesses is a two-part tariff, where you set a high entry fee to cover average cost and then price service at marginal cost. Lawyer types see this as a solution to the dead-weight loss problem, but they're wrong -- it's just another form of price discrimination. Charging higher prices to customers who value (use) the system more is just a way to increase profits.

While papers need to collect eyeballs using sensationalist headlines, the only story here is how broadband providers are price discriminating to increase profitability. File sharing is complementary to broadband access, so cable companies have no interest in hampering it. They just want their cut.

Tuesday, June 11, 2002

Lotus Notes Dealing with Lotus Notes makes me appreciate Microsoft products. No, really. A few examples: deleting emails (in Notes 5) requires marking them "to delete" and THEN refreshing your display to get rid of them. I'm not sure where they end up, there doesn't seem to be a normal trash, but deleting emails is one of the critical features that Notes doesn't support.

Replying to emails is another. If you want to quote some text, but not others, you're out of luck--it's all or nothing. It's shocking how something that offers such sophisticated functionality on some levels fails to get basic things right. By contrast I looked at the client for Jaguar (OS 10.2 beta)--auto threading, multiple accounts, built in presence, built in chat--wonderful stuff.

Monday, June 10, 2002

First day at Lotus Urk. Finished my first day at Lotus in Cambridge, MA. It's *great* to be back in Cambridge, I really like IBM, and I'm tired.

Friday, June 07, 2002

Quick links Just finished exams; I leave for Cambridge on Sunday. Here are some quick weekend links:

Hollywood is being sued to prevent it from outlawing time-shifting in PVRs. All this lawyering occurs because government doesn't have to bear the costs of its decisions.

Robert von Goeben, ex-music exec, writes that the immensly profitable record industry will never allow its content to be digitally distributed and encourages emusic startups to invest in signing on bands. I would submit that it's the inability of the recording companies to exert perfect control over their content that hampers the development of new sources of content and distribution. Free, limitless, familiar, and on mp3 is a tough combo to beat.

Late 90s bete-noir Michael Wolff says the recording industry is toast and knows it, and will fall into decline as has the book industry.

And Microsoft has come out against Hollings (Dem. So Carolina) and his insane jihad against the personal computer arguing that, well, it'd kill the personal computer.

Next week, I blog from Lotus! I am *very* excited!

Wednesday, June 05, 2002

Regulation as takings You rarely come across economics as good as this piece by Landsburg explaining why the government should bear the economic consequences of its decisions (via Jane Galt). Taxpayers do bear the costs of things like farm subsidies through higher taxes and grocery bills, but laws mandating work rules, technical specifications, zoning etc. all impose similar costs on society as taxes but are born by the regulated, not the regulator. Hollings (South Carolina-Dem, toady) SSSCA regulation would cost society billions through lost producitivity and innovation, but cost Hollywood only $287,534. With bargains like these, it's unsurprising businesses find it worthwhile to lobby.

Tuesday, June 04, 2002

Who cares about Napster for digital TV? The insane wholly owned subsidiary of Disney, Senator Hollings, has a mini-version of his dreadful CBDTPA bill edging through an industry cartel called the Broadcast Protection Discussion Group (``BPDG''). They want to ban devices that would enable digital TV programs to be shared online.

Unlike music files that are "meant" to be bought on CDs, why does it matter if free broadcast programming are redistributed by whomever? The answer has to do with pricing power against advertisers in local markets.

The big problem here is really ad-skipping a la TiVo. The broadcast industry runs on the conceit that people who watch ads then go and buy stuff AND they've convinced Congress that free-TV is a God-given right. Advertisers feel the only good audience is a national one, and since "media-diversity" nuts (and broadcasters) block non-broadcast media consolidation, the big four networks continue to corner this market (although their % viewership is falling).

So assume that ads now exist in banners at the top and bottom of the screen and TiVo can do nothing to skip them, why should the big networks care if people pass around their shows? Firstly, it shifts power from distributors to content providers (who assemble audiences by making good shows, not stringing together good lineups). Content providers don't oppose this because they need ancilliary syndication and video tape markets to make up production costs (broadcasters don't pay 100% of the cost of shows anymore).

But redistribution mostly hurts local stations, which only exist because of collusive anti-merger rules set by the government to support "media diversity" (some people actually by this, but then they join the rest of us and watch "Friends"). Redistribution shatters the local broadcast advertising markets these regional stations rely on, so they want it banned.

What makes all of this so funny is that digital TV is a non-issue. All of the above works just fine with analog signals that TiVo converts into digital with no problem. While the group claims to be talking about the future of TV, it's just about preserving the past by slipping anti-TiVo legislation in via the back door.

Monday, June 03, 2002

Is bandwidth scarce? Readers have been asking whether bandwidth is truly scarce, or whether impending limitations on WiFi are just some crazed, evil FCC plot. David Reed, Bob Frankston, and Dan Farber are big proponents, and Dan Gilmore sums up their "wireless bandwidth is limitless" position here.

Are they wrong? Well, they're certainly misleading, as the high-rated slashdot posters sussed. Here are the big sleights of hand:

1) Signals don't interfere. Signals are additive, so while they don't necessarily destroy each other, they certainly do muss each other up. If two signals are perfect mirror images, they cancel each other out completely (sin(x) + -sin(x) = 0) and while Reed's academic pedigree is excellent, the laws of physics still stand.

2) Powerful transceivers will fix it all Folks argue that mussed signals don't matter because we can stick so much software in our receivers, they can fix it all up. This is true, but it's just shifting cost from transmission to CPU (bandwidth has always been a function of transmission, processing, and storage). Since processing costs keep falling, it might make sense to reallocate resources this way, but that's a market optimization question no best left to the invisible hand.

3) WiFi is short range, so interference doesn't matter It's true that higher frequencies attenuate faster, so interference is less of a problem, but it can still be an issue in congested areas. At some range, you will always run up against interference--the question is how expensive is that to handle.

4) Mesh networks will fix it all The claim is that altruistic automatic wireless packet switching means transceivers will sort out interference by descrambling packets and passing them along, creating a giant wireless Internet. This also means users absorb the fixed cost of investing in infrastructure by buying computers with WiFi cards and marginal pricing becomes feasible. This all sounds lovely, but I remain skeptical towards spontaneous universal altruism that also effectively deals with cheaters.

I think it's unlikely the FCC will regulate WiFi spectrum--too many people own airport cards--but it'll be worth seeing 1) if networks become congested 2) how people will deal with that suggestion. Reed & Co's alturism/CPU magic sauce might work wonderfully, or WiFi might become a tragic, overgrazed common. Certainly for longer (broadcast) spectrum, the FCC should simply grant complete ownership in perpetuity to the highest bidder and let the market put it to best use.