by Ralph Nader
Address to
The Bazaar - An Open Source Software Event
New York, New York
December 15, 1999
Every time the Microsoft antitrust case moves forward, one observes a new wave of "where is the harm?" opinion articles in daily newspapers, presenting Microsoft's anticompetitive practices as harmful to competitors but not consumers.
Judge Jackson's 206 page findings of fact addressed the issue of consumer harm in ways that resonated with many computer experts. While Judge Jackson mentioned that Microsoft had considerable leeway in terms of pricing Microsoft Windows, citing an internal Microsoft memorandum comparing the benefits of a $49 or $89 price for an upgrade price for Windows 98, the findings of fact devoted considerable attention to the non-price issues, such as those relating to innovation, choice and software quality, that are key to the Microsoft case.
However, Judge Jackson's findings of fact are limited by the scope of the government lawsuit against Microsoft, both in terms of the types of anticompetitive conduct and the harm to consumers, and therefore understates the harm of Microsoft's monopoly to consumers. The US Department of Justice and State Attorney Generals have decided to prosecute a relatively narrow case against Microsoft, largely ignoring a plethora of issues relating to Microsoft's huge power in the desktop applications area, including the components of Microsoft Office, or the impact of its anticompetitive enterprise licensing strategies.
We often hear from consumers who say they are harmed by Microsoft's monopoly abuses. Here are some of the complaints. Because this is a meeting about Linux, a free operating system, I will begin with the pricing issues.
Windows is too expensive. The price for Microsoft Windows depends upon how you buy it. A license for Windows is often bundled with a new PC. That doesn't mean it is free -- only that the OEM has paid for the license.
When people talk about software prices, they sometimes forget that typically new technologies begin with high prices. Television sets, compact disk recorders and personal computers are only a few examples of this. Automobiles were very expensive when they were first introduced, costing around $10,000, or nearly $200,000 in today's dollars. More efficient mass production was followed by much lower prices. The Ford Model T, which was produced from 1908 to 1927, at one point sold for less than $260.
As prices for personal computers, scanners, printers and other computing devices have fallen, Microsoft has been able to charge high prices for many of its products. For example, the OEM prices for Windows licenses have increased, making this license an ever larger share of the cost of a new computer.
Microsoft charges consumers a list price of $109 for an upgrade of Windows 98, which is discounted by retailers to $89 -- but to get this price you must already own Windows 95, so it is like a maintenance fee. The list price for a new version of Windows 98 is $209. Yahoo.com sells Windows 98 at a discount for $181.92, nearly half the price of buying a new low end PC, and more than three times the $49.99 price for the well reviewed BeOS. BeOS is a technologically superior operating system that suffers from a paucity of third party applications, illustrating the significance of the consumer lock-in with Windows.
In addition, Microsoft is steadily tightening the conditions on licenses. Many OEM licenses for Windows are tied to a single machine, and cannot be sold or transferred to another machine, even by the original owner. Business users are facing restrictions on the use of concurrent licenses, requiring them to purchase more copies than before. And for most models of PCs that consumers buy, the OEM has to purchase the license, even if the end user doesn't want the software.
The "required to buy" Windows problem is a particular galling issue for Linux users who are often actively trying to avoid using Microsoft products. After our own efforts in 1998 to push the major OEMs to give consumers the chance to buy PCs without a Windows license, we have seem some modest improvement, as Dell and other PC manufacturers offer a limited number of PC models with Linux pre- installed. But it is still the case that nearly all of the PC models sold by major OEMs, including Dell, require purchase of a Windows license.
A consumer who has been using computers since 1995 may have already purchased a half dozen or more Windows licenses. You might have begun with Windows 95a, but bought Windows 95b so you could better use the large hard drives. And then purchased one or more upgrade computers, with new Windows licenses. Then one has to consider the number of computers that need licenses. Often a person may have separate PCs for work and home, plus a laptop for travel. So it isn't simply the price of Windows, it's the number of licenses for Windows that you end up buying, and how often you have to pay upgrade fees.
Microsoft forces upgrades of the operating system by introducing, even between official revisions, significant changes in the OS, including the important support for third party device drivers. Indeed, Windows 98 is already on its "second edition." To get what are essentially bug fixes, Microsoft charges Windows 98 users $19.95, plus shipping and handling, for the second edition of the same product. (Creating yet another opportunity to charge consumers more money so its products will function properly).
Any given version of Windows becomes obsolete within a few years, because it will no longer support the latest innovations in hardware. This is intentional, because Microsoft's biggest "competitor" in the OS market is its installed base of users who have already purchased Windows. Microsoft forces consumers to buy what is essentially the same product again and again.
In 1997, analysts said that Microsoft had a ninety-five percent share of global revenues for sales of office suites. Microsoft Office has become the global standard for word processing, spreadsheets and other desktop productivity applications. The pricing for MS Office is high. Microsoft's Office 2000 "standard" edition lists for $499, with a "street" price of $399. Even an upgrade to Office 2000 Standard has a list price of $249, and a discounted price of $195 - and this assumes you have already purchased the Microsoft Office before.
The "premium" version of MS Office is now priced at $799, or $449 for an upgrade version.
These prices are much higher than the prices for Corel's Office 200 suite, which features WordPerfect. For example, the list price for an upgrade of Corel's Standard Office 2000 suite lists for $99, about 40 percent of the Microsoft list price. (And discounts for about $79, about 40 percent of the Microsoft discounted price). Microsoft can command hefty prices for its Office Suite because consumers are often forced to upgrade - simply to read documents they receive from others. Microsoft is constantly changing document formats so that owners of older versions of Microsoft Office cannot read the newer documents. Again, Microsoft's main competitor is its own base of installed users. And, here too Microsoft is a tough adversary, using interoperability and compatibility as weapons, to force upgrades and generate more earnings for Microsoft.
Millions of computer users who have perfectly functional copies of Microsoft Office 95 found it impossible to read documents prepared in Office 97, and one anticipates a new round of compatibility issues with Office 2000.
Microsoft knows that most consumers have little use for the endless expansion of word processor features, particularly as the world has come to rely upon the much simpler formats for information used in electronic mail. Moreover, the newer versions nearly always contain new bugs, and necessitate more learning, and spark new predatory attacks on non-Microsoft products.
Plus, as MS Office and Windows become ever larger, they require huge increases in computing resources. For consumers this often means a costly and time consuming hardware upgrade -- an event highly correlated with losses of user data. But for Microsoft, a hardware upgrade is usually just another source of revenue -- as nearly every new PC ships with a new license for Windows and other Microsoft software.
One feature of Microsoft's pricing is the huge difference between its list prices and the prices paid by large buyers, including OEMs, big corporations, governments or universities. Microsoft knows that these large buyers need licenses to Microsoft products, and that they don't want to pay the high list prices. All of these large buyers get Microsoft products at significant discounts. However, for many big users, Microsoft insists on "enterprise" type licenses, which effectively force big organizations to buy licenses for many products for all employees (or students). When Microsoft gives an organization a blanket license for Windows and Office, they make it next to impossible for rivals to compete, since the organization has already paid Microsoft a license fee for all the computer users. Microsoft's pricing strategies are designed to give organizations no realistic options, if they want to avoid sky high list prices for Microsoft Office and Windows.
This is also an issue for the OEMs, since the price of software is a significant component of cost in the highly competitive PC market. Microsoft can use the threat of higher prices for OEM licenses -- for Windows or Office -- to discipline OEMs, and reduce opportunities for Microsoft competitors.
While the pricing issues are an important measure of the cost of the Microsoft monopoly, we hear more often from consumers about non- price issues, including many of the non-price issues raised by Judge Jackson.
The most common complaint is that Microsoft crashes. "At least once a day," according to many Microsoft Windows users. We also hear countless complaints that Microsoft attacks non-Microsoft products, so they don't work. For example, when Microsoft released its Windows Media player, as a competitor against the RealAudio player, consumers wrote to say it disabled dozens of third party multimedia software programs. Little wonder that people call Microsoft's Internet Explorer, the "Internet Exploder," because it attacks and disables an unpredictable number of non-Microsoft applications.
The documents in the Microsoft trial shed new light on the seemingly endless compatibility and interoperability problems with Windows and Microsoft Office. When Microsoft executives proposed making "running any other browser . . . a jolting experience," they were simply adding yet another example of the "DOS isn't done until Lotus won't run," corporate legacy.
Microsoft could never have succeeded as a software company if its intentions to sabotage third party products were known earlier, before consumers and third party developers invested billions of dollars and countless hours around the Windows platform.
Even before you consider issues surrounding deliberate hostility to users, you have the typical problem of a monopoly that can get away with poor products. Because it is so costly and difficult to migrate to a new platform, Microsoft can succeed even when its core products suffer hugely from poor stability, limited interoperabilty, and endless security problems. The fact that millions of users tolerate daily crashes of Windows says volumes about the costs of migration away from Windows.
But, as Judge Jackson points out, and as most computer experts know, not all of the quality problems are innocent. In its internal emails and by countless examples, Microsoft has demonstrated that it believes it benefits when consumers cannot make competitor's products work correctly. Microsoft has a range of methods to undermine its competitor's products. When it does not use deliberate sabotage, it can withhold important technical information or refuse to license technology to its competitors, such as when it refused to permit Netscape to distribute a utility to log-on to Internet Service Providers, or when it withholds or unexpectedly changes applications programming interfaces and data file formats.
Microsoft can also destroy the quality of rival software by using predatory business practices, such as the enterprise licensing of Windows and MS Office, exclusionary OEM and ISP licensing, or bundling of products with "must have" Windows and Office products.
When Netscape cannot effectively distribute its browser through ISP or OEM channels, and when Microsoft's Internet Explorer product is bundled in with Windows and MS Office, Netscape can no longer justify continued R&D in the product. This harms consumers who prefer Netscape. When Microsoft bundles Outlook Express, its personal information manager and email client, into Windows, millions of users who relied upon rival products, like ECCO Pro, were stranded when their products were abandoned by publishers who could not compete with a bundled product having a zero marginal cost to consumers. And there are countless other examples of this in the software market.
Despite the colossal sums of money being invested in ecommerce ventures, there is very little investment for desktop productivity software. And while the stock market seems crazy about some Linux stocks, and with all due respect to this gathering, and in light on the fact that we are using Linux extensively in our offices, Linux is still primarily a server technology, without significant penetration in the PC "client" space. For this to change, there will have to be considerable improvements in Linux documentation and in Linux desktop applications.
For most PC users, there is a steadily shrinking number of choices for a growing number of important applications. Microsoft is squeezing the life out of markets for word processors, spreadsheets, desktop database software, presentation graphics, personal information managers, email clients and Internet browsers -- the applications that most computer users need.
Some observers, such as Robert J. Samuelson, seem to think that Microsoft has provided a public service. By eliminating competitors, Microsoft gives everyone a common standard, and making life simpler has benefits, Samuelson says.
I think most people here see the poverty of this analysis. There are, of course, alternative methods of setting standards than relying upon a private monopoly. The Internet is a powerful and relevant example of how a non-monopolistic standard can facilitate enormous innovation. And, as pointed out in Judge Jackson's findings of fact, Microsoft has sought to crush third party technologies, such as Java, that create cross platform standards that Microsoft does not control.
The free software movement actively embraces a more open approach to software development. A distribution of Linux isn't the creation of a single firm. It is a collection of hundreds of programs developed by different individuals and groups, that work together. The disclosure of the source code is designed to make it easier to design software programs that work together, to solve user problems. There is competition among distributions of Linux, and users can choose alternative graphical user interfaces, programming tools, utilities and applications. As described in the so called Halloween memorandums, Microsoft's response to the popularity of Linux is to seek ways to cripple interoperability, by deploying proprietary and patented software interfaces. And so far, Microsoft has resisted efforts by OEMs to ship computers ready to dual boot Windows and Linux or Windows and BeOS.
There are, of course, huge costs associated with forcing everyone into a software monoculture. Some of the issues concern security. Microsoft's security breach of the week wouldn't be such a huge problem if its software wasn't so ubiquitous. But this is only one of many issues.
There are also large costs associated with the disappearance of the products that Microsoft crushes. In the beginning, Microsoft had a tiny presence in desktop applications, and businesses and individuals invested money and time around non-Microsoft products. The forced migration to Microsoft's Johnny-come-lately imitations is costly.
Consumers value choice, about a wide range of software characteristics. WordPerfect and Microsoft Word have different approaches to document management. Netscape Navigator, Microsoft Explorer and Opera appeal to differ users. Every software product has its own fans and its own critics. Robert Samuelson seems to think of this as an inefficiency, but the contrary is true. A "one size fits all" world harms consumers, and lowers productivity.
Competition among software products leads to innovation and improvements in software quality. This competition moves the industry to solve the problems consumers face, and leads to more productive and reliable products. Indeed, perhaps the most important consideration is that Microsoft is not a leader in product development -- it is an imitator, and this is the most significant harm to consumers -- the stifling of innovations that we never see. As pointed out by Judge Jackson, even Intel, the other half of Wintel, was forced by Microsoft to stop development of a promising new multimedia technology.
We recognize that in software markets, there may be cases where the market coalesces around a single product with a large market share. But it is one thing for that decision to be made on the basis of competition for consumer satisfaction, based upon product quality and price, and something else when consumers are forced to pick Microsoft, by an endless array of underhanded, coercive and non- meritorious tactics. Consumers are harmed when there is no real choice, except to succumb to the Microsoft Borg.
Thank you.
For more information, see http://www.cptech.org/ms
Contact Ralph Nader at ralph@essential.org