Asymmetrical Information

From the desk of Jane Galt:

I follow the Microsoft case

I follow the Microsoft case with interest, of course, given my twin interests in economics and technology. So I descended on this Scott Rosenberg article on the settlement like a ton of bricks:

However, when it comes to the heart of the problem -- the way Microsoft chose to weave Internet Explorer code into the guts of Windows in a manner that served no conceivable user benefit but that helped Microsoft elbow Netscape out of the market -- the settlement as tweaked by Kollar-Kotelly simply wimps out. "The evidence does not indicate that the removal of software code is beneficial from an economic perspective. The Court also finds that the forced removal of software code from the Windows operating system will disrupt the industry, harming both ISVs [Independent Software Vendors] and consumers."

The argument here seems to be: Let's not tamper with the innards of Windows, because who knows what might happen then? It's not an unreasonable question -- except that, by refusing to dig into the code, Kollar-Kotelly ensures that the remedies provided by the settlement will be cosmetic. And as long as Microsoft remains the untrammeled master of the Windows code, it can still find ways to trip up potential competitors without ever having to resort to anything so crude as threats to "cut off their air supply" or demands that desktop icons be removed.

Kollar-Kotelly's decision is persuasive at the level of detail; it is in sum that it fails to support its conclusion that the proposed settlement is indeed "in the public interest" -- a certification that's required by an obscure law known as the Tunney Act.

What I think this means is, even though there's a high risk that we'll screw up the company/software that formed an important part of the economic boom of the nineties, and even though we don't know how we'd actually execute this tinkering, and even though the judges who are expected to rule on the matter are in no way qualified to decide what is or is not necessary in operating system software -- despite all this, we should do it anyway.

And why should we do this? Because Microsoft is mean.

Don't get me wrong; I share the non-unreasonable fear that without some restraint, in five years we'll all be renting air from Bill Gates. But let's look at the economic benefit we're supposed to be providing to consumers by breaking the company up. Remember them? They're the reason we have antitrust law -- not to protect engineers, marketing guys, or salespeople who work at a Microsoft competitors; not to make Mac afficionadoes feel better about themselves; not to provide journalists who make in the low five figures the satisfaction of seeing a big, rich company taken down. To protect consumers from companies who use their market power to force them to pay high prices for crap.

Oh, I know. You think Microsoft software is high-priced crap. But let's look at the immediate benefit consumers would derive from having Microsoft hamstrung by Justice.

Well, in the flowering of competition that comes after their main competitor has been hogtied and forbidden to compete by the Justice department, I think that all you antitrust types can agree that Macs aren't going to get any cheaper, nor will Unix boxes get any easier to use. Those being the primary barrier to the adoption of those platforms today.

[Thank you to all those who are even now mentally composing their angry notes. Yes, my little Linux geek, I know that you find unix heart-breakingly easy to install. But I work in technology, and I've worked with the major Linux boxes, so spare me the melodrama about how it's every bit as easy and fully supported as a Microsoft box. I could get seriously hurt laughing that hard. Yes, for a reasonably adept computer user, the switch to Unix isn't that hard. But the problem with Unix people is that due to their work environment, they seem totally unaware of the larger base of consumers who cannot be described as "reasonably adept", and cannot function without those dreadful little wizards, smiley faced cartoon characters, and extensive, idiot-friendly computer help lines provided by Wintel firms.

And you, my darling Mac hound, I am well aware that your Mac is a bargain with all its video software and neato music features and the smooth, white expanse of the outer case melding seamlessly into the hip new colors on the front. . . but for the millions of consumers who don't edit their own movies or collect 8 zillion tunes on their I-Pod, all that software you consider indispensable is useless dross for which you're asking them to pay a hefty fee. I know you love your Mac. But you have to accept the fact that not everyone else wants all the features you use. And they want features you don't use, like software titles and power for a cheaper price. The ability to accept that other people have differing opinions on lifestyle questions is one of the things that makes America great. So give peace a chance.]

There might be some benefit to consumers of Mac and Unix boxes. Microsoft's development will be essentially stopped for at least a couple of years while the court decides what they're allowed to do. As anyone who's worked in software development can tell you, this is going to screw up Microsoft's software six ways from Sunday, leading to even longer delays. In the interim, some companies might be forced to start developing for other platforms, although Mac fans, don't hold your breath for them to port corporate applications onto OSX.

So net benefit: consumers of Mac and Unix boxes probably get some more software. Offset by the fact that probably more people are going to have to buy computers that are more expensive, and in the case of Unix, harder to use. We might also, while Microsoft is in its straightjacket, see a flowering of entrepreneurial spirit producing hithertofore undreamed of software products for the consumer market. Of course, we also might not. Since the benefit is fairly tenuous, we need to weigh it against the cost.

Network Effects save companies money There are advantages to having everyone on the same platform. It reduces costs for software development, training, home/work interaction, support, and makes their hardware and other software cheaper. Getting rid of the monopoly gets rid of the network effects, which will have some drag on the economy. This is particularly a propos in the case of a natural monopoly, where forcing companies to amortize fixed development costs over a lower number of sales for each OS means that the price of both operating system and other software will probably rise.

Microsoft has a lot of shareholders Kill the company, and you kill their portfolios. But we're not going to kill the company, you say? Come on. If you're having this argument with me, odds are better than even that you're some kind of a tech person. So let's take your current project. Pretend you have a little company that produces, as it's main product, whatever it is you're working on. Now imagine that the government steps in and tells you that you're going to have to change it. They'll appoint a special master to tell you what you can and can't do. You have to explain everything to the special master, wait for him to digest it and come up with a plan, and then explain to him why his plan is idiotic. Then he has to make a new plan. This isn't his only job, you know. You wait, not allowed to write a single line of code while the review process is on. A year or two down the road, he comes back and tells you what you can and can't do -- and then you have to rip apart your project and start over with the totally new specifications. Think your company would still be in business long enough to get a new product to market?

Risk is not good for the economy Getting Justice intimately involved in the operation of Microsoft raises risk in two ways. It increases the uncertainty about the future of the software/hardware market, which means companies will likely slow investment. Since falling investment is one of the primary causes of the current recession, this is a major concern. Punishing Microsoft also has a chilling effect on companies, which may fear that aggressive pursuit of market share will gain them not profits, but punishment. This is likely to deter investment and entrepreneurship, though it's hard to quantify whether this effect will be small or large. However, in our current liability environment, anything that increases the legal risk in the market has to go down in the "loss" category. Especially since the lesson that would go out from a successful drastic remedy is that competitors that lose in the market can try to win on appeal. Please don't argue that Microsoft is somehow special -- whether or not it is uniquely mean, it is not unreasonable to think that both the winners and the losers will draw this conclusion from the lawsuit, given how prominently Microsoft's competitors featured in the anti-trust suit.

Consumers like Wintel I know you don't like Wintel, because you're a hard-core afficionado. But the great mass of consumers like their Windows PC, even though it isn't as powerful as Unix, or as nifty as the Mac. They like the combination of price and features that the platform offers them. Stalling development at Microsoft for four years isn't going to make them happier, especially since

The verdict will make the main alternatives less attractive, not more. By which I mean that absent competition from Microsoft, Mac and Unix providers are less likely to improve their offerings, especially on key metrics like price and user-friendliness. Sun has no consumer marketing apparatus to speak of, and is unlikely to develop one in time, much less a software package that can be maintained by granny in Dubuque. Linux suffers from the fatal problem of all open-source, which is that no one wants to spend two years building a little animated dog to tell Grandma how to install her printer. Meanwhile, Mac's main issues -- price and power -- are unlikely to improve, first because absent Microsoft, there is no price pressure, and second of all, because there are development cycle limitations on how fast Mac will be able to put out a comparably-powered machine even if it wants to. I know I'm going to get argument on this, but c'mon -- when has getting the government to kill the main competition ever improved anyone's product?

Innovation will stall both because of increased risk, and because Microsoft won't be able to do much of anything while Justice hammers out a remedy.

So look at the modest benefits, and compare them to the major costs. I repeat that you are not allowed to consider benefits to owners of companies that might be created or get more market share, because that's not what the law is designed to do; it's supposed to protect consumers, not corporations or even their employees.

Seems to me that the benefits are all out of proportion to the costs.

Rosenberg's argument is essentially "Yes, but if we look at the costs, we can't do anything about the scourge of Microsoft!"

Mmm-hmmm. And they think free-marketers are dogmatic?

You can get rid of the pain of a stubbed toe by amputating the toe. Or you can wait for the pain to go away. The one has the benefit of "doing something about the problem", with the drawback of being, y'know, catastrophic. The other one is boring, but ultimately less painful.

As a relatively pro-market force, I get a lot of email from regulatory fans, triumphantly pointing out some place where the market has produced some result other than the one we would have gotten in an ideal world full of sweetness and light. They then wait, smirking, I imagine, for my worldview to crumble as I realize that the markets are not, after all, perfect. I think they are disappointed when I respond with the adult equivalent of "Duh. You thought we lived in an ideal world full of sweetness and light?"

Sensible proponents of the free market do not believe that markets produce the ideal result -- only that it is more likely to do so than the entity that continues to bring us the National Strategic Helium Reserve.

So why can't proponents of regulation seem to realize the same thing -- that there is not some ideal combination of regulation that will ensure that nothing bad ever happens, anywhere? I recall with fondness a discussion I had with one of my lefty readers that started with his desire to really regulate those energy companies so that California couldn't happen again, and then, as I walked him through the issues of regulatory capture and conflict of interest inherent in asking a civil service guy making 70K to watch an industry making thousands of times that amount, his tunnel vision led him into an interesting place. With each proposed solution, as I pointed out the practical difficulties that would prevent his regulatory scheme from working the way he wanted, he began proposing more and more extreme solutions to overcome the practical difficulties, until he was left with two proposals:

1) Enact the reforms he wanted even though he agreed they wouldn't work, in the hopes that at some point down the road the voters would notice and do something about it. The problem being that this is an equally strong argument for not enacting the regulation, in hopes that at some point down the road, the voters would enact the perfect scheme he desired.

2) Rip out the entire civil-service system of California so you can pay regulators enough to avoid conflicts of interest A proposal breathtaking in both its utter impracticality, and blythe disregard for cost. It would be cheaper to give everyone in California their own generator than to go through the legislative and administrative battles necessary to enact this sort of reform.

When your automatic assumption is that there is some regulation, somewhere, which can produce a perfect result, you get so focused on your goal that you begin to make ludicrous suggestions for how the rest of us should drop everything and get to work on the problem you've identified.

So don't get mad when the rest of us ignore you in favor of getting on with our lives.

Posted by Jane Galt at November 4, 2002 12:32 PM | TrackBack | Technorati inbound links