Tempering the EC's antitrust enthusiasm

In response to Microsoft's interoperability announcements last week, the EC, a force that has been pushing Microsoft down the "open" road for quite some time, had the following to say in response:The European Commission takes note of today's announcement by Microsoft of its intention to commit to a number of principles in order to promote interoperability with some of its high market share software products.

In response to Microsoft's interoperability announcements last week, the EC, a force that has been pushing Microsoft down the "open" road for quite some time, had the following to say in response:

The European Commission takes note of today's announcement by Microsoft of its intention to commit to a number of principles in order to promote interoperability with some of its high market share software products. This announcement does not relate to the question of whether or not Microsoft has been complying with EU antitrust rules in this area in the past. The Commission would welcome any move towards genuine interoperability. Nonetheless, the Commission notes that today's announcement follows at least four similar statements by Microsoft in the past on the importance of interoperability. In January 2008, the Commission initiated two formal antitrust investigations against Microsoft – one relating to interoperability, one relating to tying of separate software products (see MEMO/08/19). In the course of its ongoing interoperability investigation, the Commission will therefore verify whether Microsoft is complying with EU antitrust rules, whether the principles announced today would end any infringement were they implemented in practice, and whether or not the principles announced today are in fact implemented in practice. Today's announcement by Microsoft does not address the tying allegations.

So much for interoperability announcements acting as the magic panacea to Microsoft antitrust difficulties...though to be honest, I don't think Microsoft executives ever thought that it would make the EC magically go away. I do think, though, that despite the skepticism, IF Microsoft actually proceeds in line with its stated intentions, the commission will get a lot of what it wanted from the agreement. The EC, as it noted in its response, will investigate whether that is, in fact, the case.

My problem, however, is that the commission still feels it is necessary to have a say in what is or isn't included in a software product like Windows (the essence of the EC's "tying" complaint"). I've been a critic of antitrust enforcement for quite some time, even going so far to question whether it should exist at all. Over time, I've come to the conclusion that my problem isn't so much with government power to regulate large and powerful companies (taming market power can be quite useful), but with the way governments often choose to interpret the open-ended nature of antitrust laws.

Ensuring that dominant companies don't prohibit from a licensing or technical standpoint competitors to their products atop critical computing platforms is important and useful. I don't think Microsoft should be able to ban Firefox, via legal or technical means, from Windows, just to state a useful example. I also think they should release full and complete protocol documentation so that others might interoperate with their products. Microsoft is a company that lies at the crossroads of computing, and as such, has more obligations to be open with competitors than others.

Where I have problems, however, is when the regulators start stepping over red lines. A red line, to my mind, is a company's ability to decide what is or isn't in their product.

Regulations are regulations, some would say, and how is making product feature decisions on behalf of companies any different than saying they need to document, say, all the protocols used in their products, or saying they aren't allowed to design systems that actively prevent third-party HTML renderers (as an example)? I would say it comes down to simple and universal principles that companies can live by on their own, versus principles that are of necessity arbitrary (explained later) and require constant feedback from regulators on topics far beyond the core competency of government bureaucrats.

If I had to set it somewhat in stone, I would say the dividing line comes down to the following three principles:

  1. Whether the proposed regulation can be defined in simple terms which do not require constant feedback from regulators, allowing companies to get on with the business of creating their products.
  2. Whether the regulations actually result in better systems that benefit consumers.
  3. Whether the regulations are necessary in the first place, because let's be honest: regulations are a stab in the dark in an economic world where the most information is held by the companies actively competing in a marketplace; regulations are best kept to the minimum necessary to get the job done, so as not to weigh businesses down with half-baked solutions from sources that, of necessity, lack the same level of information as private businesses.

On the first count (item 1), I think it's all but impossible to determine what is considered a "legal" inclusion in the development platforms / end user products that masquerade as operating systems these days. Ask yourself where the proper dividing line might be. Remember, Microsoft's flagship OS didn't used to have an integrated windowing user interface, a standard TCP/IP networking stack, or even support for common concepts like threading in its earliest days. Every one of those markets either could be - or actually were - a market in which companies made money, adding extra features where Microsoft, for whatever reason, did not yet offer them.

In the antitrust trial held before Judge Jackson in the late 90s, lots of people tried to make the argument that there is no good reason for operating systems to ship with pre-included web browsers. Today, that argument seems a bit Martian, as an operating system in 2008 without a web browser is like a car without wheels. I would argue that the same applies to media playback functionality, and again, that is a product category that most operating systems (Windows included) has offered from its earliest days. The conflict only arose once Microsoft chose to enhance the functionality of these existing products to a point where they became competitive in their own right.

Paint has been included in Windows since version 1.0, yet no one points to Paint as a source of competitive concern. What if Microsoft decided to soup it up to the point where it became competitive with Photoshop? Why would it be wrong for Microsoft to do that, and how would a decision to do so be bad for consumers?

This takes us directly into the second point (item 2), which asks whether a regulation actually creates better systems that yield consumer benefits. Restrictions on the type of features Microsoft is able to integrate fails on this count. As noted, a Microsoft that decided to improve Paint doesn't harm consumers, so much as make more work for Adobe. More work is not in and of itself a bad thing, and in the Internet age, Adobe already has stacks of free competition available for download from thousands of sites across the web.

More important, I think, is recognition that there are different kinds of users in this world. A system that lacks browsers, media players and IM tools is perfectly usable to the computer intelligentsia that regularly read ZDNet, but completely unusable to those consumers who view computers as TV sets...at least from day one. For such users, pre-inclusions are like training wheels that familiarize users to what is possible.

Apple has made an entire business out of catering to consumers who want a computer that gives them everything they need out of the box -- one that is consistent across Apple's product line. That satisfies a real consumer demand, and declaring that Microsoft is not allowed to target it due to a business model oriented around sale of platform software (versus the whole package approach used by Apple) really makes no sense. Nobody would tell Toyota not to sell cars to Texas as a way of curtailing market power.

Training wheels, however, quickly come off, and that drives us right into my third point (item 3). Media Player has existed for quite some time, and though Real Networks tried to turn it into a regulatory leverage point (ignoring the fact that RealPlayer became an unstable and intrusive pile of "adware"), the fact is that Apple has already shown that the problem was never Media Player, but Real Network's inability to create a value proposition that made people want to use media playback tools not pre-included with Windows. iTunes, the media tool owned by Apple and created for use with iPod digital music players, is probably the single most popular media playback product in existence today, and it does not come with Windows.

Ditto with IM. Some form of IM might come pre-included with Windows, a place filled on the Mac platform by iChat. That hasn't stopped Skype from conquering the VoIP world, and through that dominance, pushed itself into a prominent place in text-based IM. In fact, I know several Mac users (somehow I have become a Mac support person for some of my non-technical friends, which makes ZERO sense to me) discovered the wonders of IM through iChat...and quickly downloaded Skype to chat with Scandinavian friends. So much for pre-inclusions cutting off opportunities for competitors, something which is clearly not the case in the Internet age.

Internet Explorer is still dominant in web browsers, though that dominance is quickly being eroded by Firefox. Firefox competes on quality and extra features (and I find I use Firefox more these days than I ever expected...partly because this opinionated blogger has acquired quite a preference for the pre-included spell checker). It's also worth remembering WHY IE managed to acquire the place that it did. IE 4.0, back in the day, was objectively better than Netscape by any standard you might choose to use, a fact reflected by the Mozilla project's decision to completely throw away the old codebase and start from scratch. That left the field, for all intents and purposes, to IE, which like a sporting match to which the opposing side chooses repeatedly not to show up, Microsoft won year after year.

Once Firefox started "showing up," however, things started changing, and the browser war was reignited in ways Judge Jackson believed to be impossible. The lesson from all this is: never underestimate consumers.

As I've stated multiple times before, make Microsoft document completely every protocol they use in Windows. Force them to ensure that every Microsoft-created inclusions is replaceable from a functional standpoint (e.g. always start third party media player A when an MP3 file is played, though don't make Microsoft allow defaults to be removed as defaults set the baseline of functionality). Make Microsoft allow OEMs to pre-include whatever extras third parties are willing to pay OEMs to have pre-included.

Just don't take control over what is or isn't in the base Microsoft product away from Microsoft. If regulators can confine themselves to those principles, I would be a more enthusiastic proponent of antitrust.

As they rarely can be, I am left with a rather ambivalent stance on the subject (not that the ambivalence of a Microsoft-employee-ZDNet-blogger will carry much weight with Ms. Neelie Kroes).