What if Microsoft was truly evil

Many believe that antitrust is necessary to tame the baser instincts of Microsoft. Imagine, then, a world where Microsoft is completely unfettered by antitrust concerns. The results wouldn't be as bas as many would think.

One of my favorite Marvel comics back in the day was the "What If..." series. It was a comic which considered the outcome of alternate paths characters in the Marvel universe might have taken. What if the Fantastic Four had never been exposed to cosmic rays? As it turns out, in a rather cataclysmic recasting of the theme presented in Capra's "It's a Wonderful Life," the Fantastic Four wouldn't have been around to save us from Galactus, leaving the giant humanoid free to "eat" the planet Earth (or at least the "life force" of the planet) and producing a landscape that bore a striking resemblance to Swiss cheese when viewed from space.

Galactus, eater of planets

Of course, looking back, the series was probably only interesting to a 13 year old boy who had read enough comic books as to know the tangled history of the Marvel comics universe, and thus "got" the inside jokes (except when they weren't so inside, like when they crossed "What If" with Mad magazine competitor "Cracked," which I remember thinking was one of the funniest things I'd ever read).

Asking what would have happened were Microsoft's shadow never to have crossed the desk of government bureaucrats is probably only interesting to those so obsessed with technology news that they get the "inside jokes." But then again, this is ZDNet, and you wouldn't be reading and responding to my posts if you weren't obsessed, so this is as good a forum as any.

Queuing the voice of Rod Serling....picture if you will a Microsoft left free to do whatever it wanted. I'm not just talking about a world where the EU, South Korea and the United States never filed antitrust cases against the software giant. I'm talking about a world where Microsoft never had a 1994 consent decree, which ended Microsoft's practice of granting discounts to OEMs who paid a license for every computer they shipped, even if it didn't have a copy of Windows installed. I'll go even further by postulating a world that finds antitrust so distasteful that Microsoft doesn't even have the prospect of having to face an antitrust investigation. Microsoft is simply allowed to do whatever they want, within the basic constraints of property rights.

What would happen? Economic havoc, some would say. With no limits on Microsoft, the company would be free to create a truly restrictive computing environment. OEMs would be banned outright from shipping non-Microsoft operating systems, at least if they wanted to have access to Windows licenses. Perhaps they would make it so that you couldn't run non-Microsoft software at all on Windows. Information wouldn't be available to third parties interested in developing applications that run on, or just interface with, Windows - the better to hinder competition. Penfield Jackson, the judge in the original Microsoft antitrust trial who ordered the breakup of the company (a judgement that was overturned on appeal), believed that Microsoft could potentially charge $10,000 per copy of Windows if left free to do whatever they wanted.

This extreme view, however, only makes sense in a static universe where customers can be expected to continue along historical trends in spite of actions on the part of their suppliers. Take the $10,000 / computer claim by Judge Jackson. Steve Jobs, Linus Torvalds, and fans of alternative operating systems dream of a Microsoft that chose to charge $10,000 / copy for Windows, because that would instantly inject the market for their products with so much demand that they couldn't create systems fast enough to support it.

In other words, the notion that Microsoft could charge such an outrageous fee is ridiculous, because customers would avoid paying it by choosing an alternative. Software barrier to entry be damned! Vendors would have a HUGE incentive to rewrite software currently targeted at Microsoft platforms, because the cost savings to be derived from switching would be immense.

The same line of reasoning applies to other aspects of a "super-restrictive" Microsoft. The notion that Microsoft would prevent non-Microsoft products from running on Windows can quickly be dispatched, as Microsoft would lose just as much market share as if they charged $10,000 / copy for Windows. The source of most of Microsoft's success has been in the business arena, and businesses chose Microsoft so that it could serve as the foundation of internal business systems development. That would quickly change were Microsoft to require businesses to get their all software from Microsoft.

Likewise, more money is made OUTSIDE Microsoft from the Windows ecosystem than inside it. Microsoft relies on third parties to develop the products which attract customers to their platform, and thus, they can't ban alternative products unless they want to undercut their own market share.

The same reasoning applies to Microsoft release of developer-related documentation. Preventing third parties from acquiring the information they need to make products that run on Windows would be counter-productive, and thus Microsoft wouldn't do it.

Maybe Microsoft could walk the razor's edge and ONLY release information required to write applications atop Windows, but no more, all in an attempt to stymie attempts to create products which slot into Microsoft networks. However, again, Microsoft's freedom of movement is circumscribed by what consumers want. Microsoft would have probably preferred if everyone had favored NETBIOS as a networking protocol. Unfortunately, they preferred TCP/IP, and Microsoft had no choice but to support it.

Nothing stops third parties from creating products which compete head to head with Active Directory, as an example, and a restrictive Microsoft would have a hard time foreclosing their ability to compete without closing avenues to third party product creation on Windows. For instance, the presence of Active Directory doesn't make it impossible to run a centralized LDAP directory which stores the same sort of information. Novell NDS is not blocked by the presence of Active Directory.

Some, however, want to compete with Microsoft software from the inside, by creating products that look, for all intents and purposes, like a Microsoft product.

This is what the SAMBA folks have done, however, and though they have had to reverse engineer a certain amount of technology, they have been surprisingly successful at doing that. Or rather, maybe it isn't so surprising, as keeping secrets in software is hard to do so long as someone is willing to take the time to untangle the protocols and technologies used. Microsoft COULD try to shift the protocol definitions as a way to make their products a moving target, but that carries costs to Microsoft as much as it does to SAMBA developers.

Last, what about Microsoft restrictons on OEMs? Microsoft prior to 1994 did provide volume discounts contingent on charging a fee for every box shipped by an OEM, whether or not a copy of Windows was installed.

However, I question the real value to Microsoft of such restrictions. It's worth noting that most large OEMs STILL don't ship PCs without a copy of Windows pre-installed, even though government oversight has prevented Microsoft from charging higher fees should OEMs opt to ship with a non-Microsoft operating system.

The fact of the matter is that large vendors of personal computers are really large vendors of WINDOWS personal computers who have chosen to achieve economics of scale by standardizing on the provision of systems that meet the needs of 95% of potential customers. These are customers who have no real desire to offer systems with alternative operating systems, which makes restrictions designed to ensure they have this choice about as effective as offering a version of Windows with Media Player (but not the media pipeline, on the request of Real Networks) stripped out.

Alternative operating system vendors would have as much luck convincing Windows OEMs to ship with their products as sailing fans will have at convincing Toyota to make sailboats. A more productive effort at distribution would be through smaller vendors that specialize in the provision of systems that do not run a Microsoft operating system. Even in an alternate universe where Microsoft threatens to kidnap Michael Dell's dog should the company's founder look cross-eyed at a copy of RedHat Linux, Microsoft wouldn't be able to do anything about small vendors who have no need of bulk licensing agreements with Microsoft.

These small companies have about as much incentive to try to compete on price with the large Windows OEMs as I do to create a car company that competes with Toyota. Even if they licensed Windows, they couldn't get anything near the price the large bulk-sellers of Windows PCs can get. This reduces the cost differential between shrink-wrapped Windows (which would remain low in cost for the reasons I described earlier) and the OEM-priced Windows, at least for the little guys. Given that the Windows PC market is occupied by huge companies with massive hardware and software economics of scale (a cost that matters more to OEM success than the differential between shrink-wrapped Windows license fee and the bulk sales fee), smaller players have a large incentive to go after the areas of the market with less competition.

The only way to really change the situation vis a vis large OEMs would be to FORCE them to offer non-Microsoft software, which besides being a rather odd solution to Microsoft's supposed "predatory" ways (Microsoft too powerful? Put restrictions on the purchasing habits of Microsoft's customers, which is like a Kurt Vonnegut plot line), is also a recipe for raising costs on OEMs who got to where they are by choosing to standardize on a particular hardware and software offering.

Though I'm stating the obvious, I'm skeptical of the need for antitrust. Though my "What if" parable describes one reason, here's another, one that's related to the case that inaugurated American enforcement of antitrust and served as the case which cemented within the American consciousness the supposed importance of the Sherman antitrust law of 1890.

The breakup of Standard Oil occurred against a backdrop of a company facing a technology inflection point. Standard Oil's "monopoly" was in kerosene, a market for which they had a 90% share in the United States.

Petroleum, however, was the market of the future, and Standard did NOT have such a dominant share of that market. At the time of the trial, Standard's share of that market had fallen to 64% as they faced competition from large, well-capitalized competitors such as Texaco, Gulf Oil and Sun. New sources of oil were being discovered every day, and though this has no bearing on market competition, many of those sources of oil would be nationalized in a few decades by Arab governments looking to take control of oil production, placing them well beyond the control of American capitalists such as John D. Rockefeller.

To quote Dominick Armentano, a professor of economics athe the University of Hartford who opposes antitrust as national economic policy:

Should people be concerned about monopolies? Of course they should. But we must understand the true source and causes of monopolies — governmental barriers to free and open competition. The solution to the monopoly problem, then, lies not in antitrust laws (which should be repealed) but in the repeal of all governmental barriers to free and open trade.

It's an interesting theory, and one that makes a certain amount of sense to me, particularly given that Mr. Rockefeller often used government influence as a means to fend off competitors (something the AT&T of old was famous for doing as well). That being said, I do see an area where government influence could be useful (or at least less bad compared to current antitrust policies)...provided antitrust regulators understood the concept of restraint and could resist the impulse to attempt to micromanage the marketplace. But, more on that (hopefully) next week.