Last week, I looked at how Microsoft wound up facing monopoly and antitrust complaints from friends and enemies alike. Now it’s time to see if the charges are relevant or leftovers from a different economic time – and why only Microsoft seems to be facing such scrutiny.
Could Microsoft Learn From Apple? Why doesn’t Apple get complaints like those against Microsoft? A former Mac OS clone vendor has filed suit against Apple, claiming that Apple abused a monopoly position in Mac OS hardware to kill clones in 1997 – but no one has filed a similar suit claiming Apple has abused a software monopoly. It’s surprising, because software firms love to file lawsuits. It’s a status thing.
Apple generally avoids such difficulties by building systems that invite third-party participation and not tightly integrated components that can’t be separated. Think about your Internet software. You can’t rip the email functionality out of Netscape Communicator and put Eudora or Mailsmith into your browser, but you can use Eudora or Mailsmith instead of the built-in email capability. That’s the difference.
Look at AppleScript. It’s based on Apple events – an inter-application communication mechanism useful in many ways other than scripting. Apple events in turn are based on other Mac OS components like low-level events and AppleTalk. None of these technologies is restricted to AppleScript – developers are free to use any of them without including AppleScript support, or even if AppleScript isn’t installed.
But Apple didn’t limit the Mac OS to a single scripting implementation. UserLand Frontier was available before AppleScript hit the market, and programs from CE Software’s QuicKeys to MacPerl to WestCode’s OneClick all provide some level of system-wide automation. Anticipating third-party interest, Apple constructed a layer called the Open Scripting Architecture, or OSA. (In fact, the technical name of a scripting addition is "Open Scripting Architecture Extension," or OSAX.) The mechanisms for scripting, but not the actual AppleScript language, were made available in the system to any scripting language. Through this layer, all Macintosh scripting languages can work together if they want. That’s why the Script Editor lets you edit and compile scripts in any OSA-compliant language you have installed (clicking the word AppleScript in the lower-left corner of a Script Editor window activates a pop-up menu for selecting languages).
AppleScript comes free with the Mac OS, but other scripting languages do a better job of meeting other needs. This kind of open-ended support is typical of the Mac OS. You may not know it, but even fonts are now handled this way. QuickDraw GX introduced the Open Font Architecture, which lets developers insert their own code to draw fonts in their own formats. When OFA is available, Adobe Type Manager works as an OFA font scaler, as does the built-in TrueType scaler. OFA is available without QuickDraw GX in Mac OS 8.5 and later, though documentation is still not available to mortals.
If Microsoft had created some kind of "Open HTML Rendering Architecture" and supplied Internet Explorer’s engines as one implementation, there wouldn’t be much problem. Netscape could write their own components to replace Microsoft’s components, and customers would have their choice of code to use underneath the application level. For instance, Eudora Pro 4 for Windows would ask the operating system to provide HTML functionality, and users could choose whether that functionality came from Microsoft, Netscape, Spyglass, or from any other developer of HTML renderers.
If Microsoft had done this, it’s unlikely the Justice Department would have had a complaint concerning bundling. If the government were more engineering-savvy, this might be the kind of remedy they want a court to impose if they win their suit. Such a solution addresses problems of Microsoft stifling competitiveness without swinging the pendulum too far the other way – forcing Microsoft to bundle Netscape’s code when the two have legitimate competitive issues about middleware layers.
Yet it’s the absence of this solution that, to me, shows more of Microsoft’s probable intent. Microsoft hasn’t been shy about appropriating Apple ideas in the past, and they’re currently in the process of lifting the scripting architecture for use in Windows. Microsoft knows how to do the third-party expandability drill, as they’ve shown in areas like ActiveX. By deliberately ignoring these well-known and successful ways of allowing third-party integration – and bundling Internet Explorer and only Internet Explorer with Windows 98 – Microsoft has inadvertently shown just how intent they are on controlling the middleware layer and not just providing the new features they’re constantly crowing about.
I see little doubt that Microsoft broke at least the spirit and probably the letter of the antitrust laws. Internal Microsoft memos released by the U.S. government have shown Microsoft trying to drive Netscape out of business, using Windows as a club wherever they believe it will work. If the Justice Department establishes these allegations as fact in Judge Jackson’s eyes, the government will have made its case. Using Windows to muscle other software into everyone’s hands is exactly the kind of behavior the antitrust laws were designed to prevent, or at least to punish.
Does Microsoft’s Behavior Matter? Are these alleged violations of U.S. law something that means Microsoft should change, or something that means the laws should change? After all, these laws were written in the days of an industrial economy. When a company in 1912 included something "free" with another product, the free item had a definite cost. It’s fair to guess that none of those legislative authors ever conceived that the entire cost of a product could one day be tied up in research and development, with virtually no cost for each individual unit. Once Microsoft creates Internet Explorer and makes it available for download, the "cost" of each new user is only the barely measurable cost of the bandwidth for the download. Even the classic information distribution methods of newspapers and books have production costs. How do – and should – the government and citizens respond to monopolistic behavior in the information age?
Microsoft argues that their behavior is competitive, not anti-competitive. In a 1998 Wall Street Journal op-ed piece, Bill Gates says the main reason Microsoft is a target of government investigators is that they innovate too well.
"The government’s proposals for how software products should be designed would hurt Microsoft, but they would hurt others more. They would deny independent software developers the ability to make use of the latest operating system technology in their own products. And they would deny consumers the ability to buy innovative software that allows them, say, to download data from the Internet while they are using Microsoft Word or Intuit’s Quicken. Some of our competitors don’t think consumers should be allowed to benefit from such innovations. They don’t want to compete in the marketplace; they want to compete in court."
As usual, Gates is twisting the truth in his desire to win. Nothing stops anyone from downloading material from the Internet while using Word or Quicken today; browser functionality doesn’t have to be built into the operating system for Windows’s vaunted preemptive multitasking to work. But the real battle here, as an article in Information Week points out, is more about bundling and architecture than anything else.
Whenever Microsoft releases a new product, part of the feature list includes new code added to address customer concerns (or Microsoft’s strategic concerns), and qualifies as "new features" by anyone’s definition. But part of the feature list almost always includes formerly separate products that Microsoft tosses in for free. If a particular expensive stand-alone server isn’t doing well, Microsoft may decide to make it free with another, more popular server, in the name of adding functionality to the more popular product.
The trouble here, of course, is that Microsoft is essentially giving away certain products to build market share – the same products can’t win in the free marketplace. Information Week talks about a new Microsoft Online Analytical Processing (OLAP) server bundled with the latest release of Microsoft’s SQL Server (version 7.0). Existing OLAP servers, reports IW, sell for up to $3,500 per user, but Microsoft’s will be free with their popular SQL Server. If Gates’s mantra about "choice for consumers" were more than an empty marketing slogan, his company wouldn’t routinely undertake moves like this to eliminate those pesky competitors.
He could be the world’s most self-absorbed person, but Oracle CEO Larry Ellison may have been on target about Microsoft. Ellison may have studied Microsoft more closely than anyone else in the world, since his overwhelming drive in life is to be Bill Gates Except With Style. Ellison thinks Gates’s mantra of "choice" via bundling is, to no one’s surprise, total crap. The Oracle chairman has told interviewers that Gates’s logic sounds good, except that what he sees as "innovation" is whatever his major competitor’s main product is. What if Oracle becomes Microsoft’s next strategic threat? Will SQL Server become a free part of Windows NT – or Windows 98? Will Gates then argue that he can’t "innovate" without the ability to take any piece of application software and build it into the operating system?
Microsoft’s strength as a combined application and OS vendor has always been integration, like the simultaneous release of Office 95 and Windows 95. If Microsoft’s servers are the most integrated with Microsoft’s operating system features, it follows that Microsoft’s servers already enjoy a serious competitive advantage. Tight integration, as with Internet Explorer, makes it impossible for competitors to replace functionality. If Microsoft then uses that position to destroy competition by bundling inferior servers, giving away a mediocre but free built-in product in a high-quality but high-price market, then in the end they’re using the monopoly OS position in more than just the Internet market.
What’s the Result? Most roads lead to the same conclusion: Microsoft’s behavior is bad for the industry and thus bad for consumers. It’s not about giving people more choice; it’s about giving them the choice of Microsoft for free or other products for way more money. Once the choice to use Microsoft products would be too expensive to change, Microsoft can then start raising fees and moving their servers towards working well only with their applications, locking in more Microsoft choices. Think about how many offices today have already standardized on Microsoft Windows and Microsoft Office. Will any situation be helped by those same offices locking into Microsoft Internet software, Microsoft server software, and who knows what else?
When a company obtains a market share as large as Microsoft’s, the rules simply are different. A small company can use a strong product to wedge its way into new markets, just like BMW has done by using its niche in luxury cars to enter the upscale sport utility vehicle market in the U.S. A large company can’t use this sort of tactic, because the lack of competition in one market gives them an unfair advantage over the others. Even if other makers of OLAP servers did decide to give away their products (and their revenue), they couldn’t reach all the customers of Microsoft’s SQL Server. In short, anti-competitive actions may be unpleasant and even unethical when practiced by small companies, but they don’t violate antitrust law, as they do when they’re practiced by monopolies.
If Microsoft hadn’t shown signs of anti-competitive behavior in the past, maybe its critics could cut the company some slack. If Microsoft’s version of "innovation" wasn’t so often just buying or copying a product and giving it away at lower prices, the arguments might not ring hollow. If Microsoft’s examples of competition were companies that had thrived instead of barely hanging on, they might sway detractors. But they have, it has, and they are, and it all adds up to trouble. Messy though it is, and unsympathetic though U.S. courts sometimes are to antitrust claims, the Department of Justice is moving wisely to at least keep Microsoft in check.
The recent past has shown U.S. citizens that ignoring antitrust laws doesn’t lead to more choice – it leads to larger companies with fewer options. It’s become apparent with banks, cable TV providers, regional phone companies, and others too obscure to mention. It may not be the government’s business to regulate Adam Smith’s Invisible Hand, but the government does serve the people by preventing too much power from being vested in non-elected bodies. The strange part is that Adam Smith agreed – he was strongly opposed to monopolies and corporations.
The U.S. Great Depression of the 1930s showed us how dangerous large and unrestricted banks can become – not because they don’t operate well, but because the consequences of failure are too much to bear. The 1970s and 1980s showed that when large companies would seriously damage the U.S. economy by their failure (Continental Illinois Bank, Chrysler, and recently Long-Term Capital Management), those companies are bailed out instead of letting the Invisible Hand run them through. The result could easily be a Microsoft that has freedom to forget everything it knows about innovation and competition and still remain the strongest software company in the world – at taxpayer expense.
Microsoft seems beyond failure now, but if their products become so good that upgrade sales evaporate, the story could change. No company is crash-proof, and no interest other than Microsoft’s is served by so much power residing in one company’s hands. If it really was about innovation and choice, no one would care.
Implications for Apple — Had things gone differently for Apple, this might be a battle fought in Cupertino instead of in Redmond. Pundits love to speculate that Apple would be in Microsoft’s position had the company licensed the Mac OS in 1985 like Bill Gates wanted. Unfortunately for revisionists, there remains no evidence to support this conclusion, making it speculative at best, sensationalistic at worst.
But Apple has already seen a taste of this, as noted earlier, with frustrated cloners claiming Apple is in the same position as Microsoft, just over the smaller Mac OS market as opposed to the entire PC market. (Microsoft is trying to avoid prosecution using a similar tactic, claiming they don’t have a monopoly on all computer operating systems, especially if Unix is added to the mix.) If the claims against Microsoft succeed, Apple could find itself in similar straits by those who want to force the company to restart a cloning policy. If done under court order, cloners would be sure Apple wouldn’t suddenly decide to end all the contracts again, and Apple could once again be in financial trouble.
While breaking Microsoft’s stranglehold on competition bodes well for the PC industry, the major effect these specific lawsuits would have on Apple is the strengthening of Netscape. Microsoft argues this is the only reason they’re being persecuted, but Netscape is simply the strongest competitor in the market Microsoft is trying to corner. The strongest second-place company will always benefit when a monopoly is broken; this one just happens to be Netscape. A stronger Netscape could help the Macintosh browser market, although the Mac OS hasn’t been Netscape’s platform of choice for some time. And Netscape isn’t above this kind of behavior, either – they announced an OpenDoc part for their browser in 1996 and never followed through, largely to prevent investment in Spyglass’s set of OpenDoc Internet parts. It’s good to remember that hypocrisy is typical in business allegations.
If the Justice Department succeeds in making Microsoft create a more open operating system, Apple may reap benefits. When developers and users aren’t so locked into Microsoft technology, the Macintosh becomes a more viable platform. Such folks wouldn’t depend on Microsoft largesse in implementing Microsoft technologies for the Mac OS just to render it suitable for their use, and that won’t hurt a bit.
There are also the obvious benefits – testimony in the trial so far has alleged that Microsoft tried to make QuickTime for Windows look buggy and threatened to discontinue Microsoft Office for Macintosh unless Apple chose Microsoft Internet Explorer as the default Web browser. However, the plaintiffs are not alleging that Microsoft has a monopoly in application software. The plaintiffs’ introduction of these claims – which are not directly related to the Windows monopoly – is risky. If they prevail, the judge may grant them broad remedies; if they fail, Microsoft will have been "cleared" to proceed with these business practices. And that’s the rub.
Remember what Microsoft did after Apple’s interface lawsuit of 1988? Microsoft prevailed in the infringement battle, largely because Apple unwisely and unwittingly licensed Microsoft to use interface elements. Once the U.S. Supreme Court turned down Apple’s final appeal, Microsoft wasted no time in appropriating more Macintosh user interface elements for Windows 95, and even more in Windows 98. If Microsoft wins this battle with the Justice Department, the kind of behavior we saw from them before the investigations started could pale in comparison to the ensuing bundling, integrating, and wholesale elimination of competing products. The stated goal of Microsoft is to see a computer on every desk on the planet, and Microsoft software running on all of them. That’s not a goal you achieve by being either shy or scared of lawsuits you’ve lost.
You may disagree, and that’s great. Your ideas may be better, and only through discussion and debate will those come to light and implementation. But no matter how Microsoft tries to reassure people, I still think it’s a problem. A monopoly position almost always is, because power still corrupts. Unlimited economic power in the hands of the richest man in the world can spell real trouble for anyone who decides to compete with him, as lots of companies have already seen. The Invisible Hand has difficulty controlling a market where individual responsibility is divorced from corporate interests. Even Adam Smith knew this, and what the world has seen of such experiments so far isn’t encouraging. Given Microsoft’s stated goals, keeping the powers in check remains a good idea.
[Matt Deatherage is the publisher of MWJ, an acclaimed subscription-only newsletter for serious Macintosh users. Those who sign up before 01-Dec-98 for a free three-issue trial subscription can still receive MWJ’s Mac OS 8.5 special edition, the most comprehensive coverage of Mac OS 8.5 available anywhere.]