© The Financial Times Ltd 2016
FT and 'Financial Times' are trademarks of The Financial Times Ltd.
The Financial Times and its journalism are subject to a self-regulation regime under the FT Editorial Code of Practice.
Last updated: September 20, 2007 4:03 pm
A top European Union court has rejected Microsoft’s appeal against a landmark antitrust ruling, dealing a serious blow to the US software giant in its nine-year battle with European regulators.
The original ruling by the European Commission that found the software group had violated competition rules by abusing its dominant market position. The record €497m fine imposed by the Commission in 2004 was also upheld and more fines are possible.
Is this latest development good news for consumers and rival businesses, showing that regulators have real power? Should further probes against Microsoft, perhaps over new functions bundled into Vista, and other dominant groups now proceed? Or will the action actually cause harm by distorting the market for software?
Our expert panel includes Professor Harry First, at New York University School of Law, and Richard Epstein, professor of law at the University of Chicago. Professor First has served as Chief of the Antitrust Bureau in the Office of the New York State Attorney General. Professor Epstein argued in March that “the cynics, who have often denounced competition policy as a covert weapon for undermining market competition, are the only ones who should rejoice at the ham-handed protectionist manoeuvres of the EU”.
The panel answered your questions live at 2pm BST (9am EST) on Thursday September 20.
Will this ruling have any tangible effect for European (or even US) consumers - or is this merely a high level power struggle?
Alan Hardacre, Brussels
Richard Epstein: There is no question that this is a power struggle because it is hard to see how the Court of First Instance would have applied the same rules to domestic EU corporations. There is both the natural home court bias, and the brute fact that most of the dominant firms are American, which is why the follow arguments are addressed to the likes of Google and Intel.
But it would be a mistake to assume that it is only a power struggle. Large fines influence incentives. One obvious implication is that there is more likely to be delays in innovation as companies ponder the antitrust or competition policy implications of what is done. And if the penalties levied are too high relative to any conceivable offense, then overdeterrence is a real issue, as is surely the case with the Media Player side of the case.
Harry First: I don’t think the ruling is likely to have much immediate impact on consumers. One of the European Commission’s remedies for Microsoft’s conduct—requiring Microsoft to sell a version of Windows without the Media Player—has not been a popular product in the marketplace because it sells as the same price as Windows with the Media Player. The other remedy—licensing certain information to competing software developers—has yet to be implemented and may not prove very helpful to software developers anyway. The future effect, though, may be to make it somewhat easier for small innovators to compete with dominant firms by forbidding dominant firms from taking actions that exclude them from the market. This could mean more innovation that dominant firms can’t control.
Does Microsoft use monopoly power to push out the competition or is the production of an OS just so far out of reach for most companies that no-one else has the money or expertise to compete? Can Google - in the form of their office software - and Apple and Linux really be competition for Microsoft?
Julian Mitschke, Schoeneck, Germany
Harry First: Microsoft has engaged in a long campaign to keep competitors from emerging that might challenge its position in the OS market. This effort — particularly as it was directed against Netscape and Sun’s Java technologies — was well-documented in the US litigation against Microsoft.
It’s not just a question of money or expertise. Google and Apple offer different sorts of competitive challenges to Microsoft. At this point, neither of those challenges affects Microsoft’s position in the OS market. Linux is a direct competitor in the OS market, but its efforts have not had great impact, as is shown by its small market share. I don’t think that anyone really expects Microsoft’s dominant position in the OS market to change in the near future.
Richard Epstein: I think that the reason Microsoft has a huge leg up on everyone else is that it has poured more into this venture than anyone else. To say that it has pushed others out of the business does not give us any information as to whether the means it has used for its success are proper or improper. And the EU did not stop to ask, as it should have, what the source of dominance was. Nor did it stop to ask whether dominance was a per se bad.
If in fact it is the result of superior imagination and strategy, the only losers are the competitors, not the customers of the firm. That is one reason why the US Department of Justice, which signed a very different consent decree with Microsoft was so exercised over the case. It thought that the definitions of interoperability in this case where too broad relative to the the risks at hand.
Do you feel the decision has been decided on the correct basis, that of protecting competition, or has it been decided on more of a political basis with the court’s arm being forced by the Commission’s determination to assert their authority?
Brett Ducker, University of Leicester
Richard Epstein: There is no doubt that the CFI has some incentive to back the commission, for the home court advantage. That comes up in the usual ways for appellate review. Certain factual questions are resolved in accordance with the Commission’s findings, as is customary on these matters. And it was quite clear that the CFI also deferred to the expertise of the Commission on the points of law.
What exactly is going on is hard to tell because the style of the EU opinions are highly opaque. There are no individual authors to these opinions, no concurrences and dissents; and the discussion of case law reads in a ponderous and bureaucratic way, which makes it hard to discern real motivations. The statements by Neelie Kroes after the case were much more in the bare-knuckle tradition. The animosity toward Microsoft was palpable. But nothing remotely like that could be found in the CFI’s decision.
Harry First: The Court of First Instance has hardly been a rubber stamp for the European Commission. In a number of recent decisions, in fact, it has rejected the Commission’s decisions and criticized its work. The Court’s very thorough review of the facts of the case and the parties’ arguments gives the impression of a careful effort to test what the Commission did to determine whether it followed the requirements set out in the prior case law. I didn’t get any sense of a political agenda (although the opinion does convey some sense of exasperation with some of Microsoft’s arguments).
Does this ruling serve more the needs of Microsoft’s competitors than the consumers?
Harry First: I think that the intended beneficiaries of the ruling are consumers. The goal of the Commission’s effort was to provide more consumer choice, both in server operating systems and in media players, and to make sure that Microsoft didn’t control those choices through its dominant position in the desktop operating system market. The way to give consumers those choices is to make sure that there are competitors around that are able to offer the choices. So, yes, the ruling (if it makes a difference) will serve the needs of those competitors, but only because we expect that competition will force those competitors to serve consumers’ needs.
Richard Epstein: One of the true difficulties with the EU decision is that it did not articulate the major premise of its own work. The words abuse of dominant position rolled off its tongue with great ease, but the translation from that generality to some proposition about consumer welfare requires one to take into account the pluses and the minuses of the various practices.
Here the key point to remember is that the Microsoft position has evolved substantially since 1998 when this complaint was first brought. In its earlier incarnation the firm took the view that it was just another firm entitled to complete autonomy in the way in which it sold and priced its goods and services. The network part of the equation, and the potential duties to deal, were not part of its thinking. But as the American law suit evolved, Microsoft gave up this first, absolutist, line of defense and moved to sounder arguments that recognized the need for other suppliers of applications to have access to the Microsoft operating system in ways that guaranteed reliability and interconnectivity.
The EU decision clearly demanded more than this of Microsoft for its definition of interoperability was far broader, and involves at the very least the duty to surrender key trade secrets about the internal operation of the server which will make it easier for rivals, without incurring expenses, to duplicate the Microsoft server instead of just connecting to it. That is a huge shift because it raises the serious risk of cross subsidies running in only one direction. And if my innovations can be captured by you, without payment of equal to the loss revenues that I sustain, then there is a drag on innovation.
So put it all together, and the cross subsidy helps the competitors and hurts the consumers. It does not matter in my view that the source code need not be turned over. So long as the information that is turned over makes it easier for others to replicate what Microsoft has done without its research, then there is a serious risk to consumer welfare.
When I was young, automobiles rarely provided air conditioning. Those who wanted it, would buy a unit from Sears Roebuck and install it. But the unit was clumsy, and got in the way. Now, almost every car has air conditioning built in. It is more aesthetically pleasing, more likely to be matched to the power of the engine, and doesn’t obstruct leg room. So, Microsoft has added a feature to its operating system, how is that significantly different from other engineering upgrades?
Alexander Miczo, Mt. View, CA
Richard Epstein: The short answer is that it isn’t different in terms of its efficiency gains, as noted in the answer to Samuel above. The argument of the EU is that the case differs from the auto situation because of the dominant position that Microsoft has on the operating system, which means that its various players are the only ones that have this advantage. So the question is why is this bad?
One argument is that there is some exclusion here of others. But the exclusion is quite weak in this case, because it is now agreed that others should be able to know enough about MS code to make their software compatible with it. Downloading is very efficient so that the losses involved are neutralized in part. In addition, the integrated product, to the extent that it does yield higher returns, is likely to accelerate platform improvements from which others share.
There is an argument that distribution is the key feature in this case, and that Microsoft should be forced to make sure that its programs are desired by offering them only on the net. Note that the EU did not require that Microsoft sell off its new systems without the media player. It only required that it sell a version without it, which naturally enough no one wants to buy. My guess is that consumers like both features: built-in compatibility and downloadable, as it were. It is simply silly to treat the failure to offer the stripped down version that no one wants as an abuse of a dominant position. Either you stop the integration or let matters rest. Consumers prefer the first, and the EU Commission is reluctant to buck them, for then there would be a real reaction.
Harry First: You have raised a really good question: when do separate products become a single product? Your air conditioning example is excellent (I confess that I have used it in my classes!), but one could multiply it when thinking of all the things in automobiles that started out as add-ons and that we now think of as part of a car.
Same is true for software — in early word processing programs, footnoting was a separate product. On the other hand, Microsoft has not added Word into Windows. Could it do so and simply call it a ”feature” of its operating system? Frankly, I’m not sure I have a perfect answer for all products for all time.
What the courts have tried to do, I think, is to see whether the efficiencies of integration are so strong that it simply makes no economic sense to offer the two parts separately. This sometimes calls for a judgment that looks at historical practice (is there a separate consumer demand for the two products? are the products now offered to consumers separately? were they ”born” separately or were they always integrated?) and at the reasons for integration advanced by the defendant.
With the Media Player, both favored a finding of two products: Microsoft itself offered the Media Player separately, as did unintegrated competitors, and there was not a strong technical case for offering the product only bundled with Windows (in other words, it was not inefficient to unbundle as well as bundle). In the end, of course, the hope is that the gains to consumers from having competition in this product (here, innovation and consumer choice) will outweigh the costs of unbundling.
The interoperability will enforce competition and therefore lower prices. But will we get lost in the load of different software and programs that will appear in the future? Isn’t bundling useful?
Samuel Urech, Chur, Switzerland
Harry First: There are two different aspects of the rulings, although both are intended to increase consumer choice. The interoperability information part will enable sellers of competing server software to compete more effectively with Microsoft by insuring interoperability among servers and with desktop PCs.
The unbundling of the Media Player is intended to give consumers the ability to have Windows with the media player of their choice and without Windows Media Player. There is, of course, some cost to having choices, in the sense that consumers will want to try to understand their options, which is sometimes a difficult task. Still, we generally prefer that alternative to having a single seller determine what it is that we can buy. And you are certainly correct that bundling is useful (it cuts down the cost of making all those choices); it’s why we see bundling throughout the economy.
The Commission’s decision, though, didn’t forbid bundling. Microsoft can still offer a bundled version of Windows and the Media Player. What the Commission did was to make Microsoft give consumers a choice between the bundled version and the unbundled one.
Richard Epstein: The interoperability will lower costs in the short run whether or not it ”enforces competition” or produces cross subsidies. But the consequences will not be the same in the long run. If there is an enforcement of competition, then we have reason to expect that prices will remain lower in the long run.
But if the short-term lower prices come from an appropriation of intellectual property, then in the long run prices will rise to consumers, as innovators will slow down to take into account that some portion of their gain will belong to others. Allowing people to take land for farming after it has been cleared by its owners lowers price in the long run as well. But the land will not be cleared and fertilized to the same level in the long run if the practice continues.
As to bundling, the answer is that it does produce net gains, which go in this discussion under the name of ”integrated products.” By making tight connections, Microsoft has better coordination across applications, which is an efficiency gain, but one that received scant recognition in the CFI.
On the assumption that there are no natural barriers to the degree of influence which can be exerted by uncontrolled extensions of monopoly, would you agree that the continued existence of markets, and the relevance of the capitalist system itself, is dependent on the ’market space’ protected by effective antitrust enforcement?
Nick Hill, London
Harry First: I think that, at some point, most monopolists will do themselves in and their power will diminish. That’s because market forces will eventually work. The question is, when will that point come and what’s the cost of waiting. You are correct, in my view, in saying that we need antitrust to be sure that those market forces work sooner and that society doesn’t pay the costs of long-run monopoly.
Richard Epstein: I think that the answer to this question is the oppose. There are powerful limitations on the ability of anyone to maintain a monopoly on a global market, or to extend any monopoly they have. The wider the market, the harder it is to keep out rivals from one or more niches. In the end exclusionary tactics fall to superior technologies. There is a lot to be said for a sensible antitrust policy, which is directed chiefly toward cartels.
But the logic of antitrust in the area of unilateral conduct is much harder because the efficiency component of the relevant conduct is far greater. And the weakness of the EU CFI is that it treated Microsoft as engaging in pure abuse, forgetting that it has provided the platform on which countless other advances by all sorts of licensees and third parties has taken place.
Does this verdict have any effect on the settlement that Microsoft entered with the US authorities? Principally, can parties (like the Attorneys General of several states and users like me) who suffered from Microsoft abuse go back for additional redress?
Neil Marathe, Doha, Qatar
Harry First: The decision of the CFI will not likely affect the US cases. The district court that entered the settlement decree (and a decree in the companion litigation continued by several of the states) is now considering whether to extend that decree, which is mostly set to expire in November. I don’t think her decision will be affected by the European case because the European case involves different facts and legal theories (although some of the relief ordered by the European Commission is similar to the relief on the US side). As for whether you can sue, courts in the US have held that non-US citizens who purchased Windows outside the US cannot sue for damages under US law. Sorry!
Richard Epstein: On that question, there is no real doubt. A settlement is a final disposition of the case, and it cannot be opened again, even if there was some abuse. I disagree with that conclusion. You should know that I have worked for Microsoft, but you can read my arguments about this in the AEI book Antitrust Consent Decrees in Theory and Practice: Why More is Less (American Enterprise Institute 2007), which defends the position taken by the District of Columbia Court of Appeals, which ratified the settlement and rejected all the claims of the dissident states. The US Department of Justice was not amused by the CFI decision and attacked it in what is surely an unusual maneuver. But the US case is closed, and will remain such. The key issue going forward is that of coordination and compatibility between our two antitrust enforcers. There is a large gulf today.
Presuming that the rule of law is a good thing, doesn’t it follow that companies must be punished when they are found to have broken the law?
Brad Stiritz, Chicago
Richard Epstein: That question is quite the mouthful. Yes parties should be punished because they have broken the law. But that proposition does not solve the problems at hand.
The rule of law has other elements as well, one of which is that the requirements of the law be knowable so that people can conform their conduct to it. That is far from obvious in this case, as the requirements for dealing under the abuse of dominant position are, to put it kindly, opaque in these settings.
Taken to its limit, the refusal of a dominant firm to give the competitor whatever it asks for should not be regarded as an abuse of a dominant position. So there has to be some limiting proposition as to what can be demanded, and the EU CFI did little to define the scope of the obligation. Next, the punishment must fit the crime because over-deterrence is risky as well.
Clearly the response to the media player part of the case was over the top. Suppose that there is a duty to supply the stripped-down version that no one wants. Keeping it off the market is a wrongful act with no real consequences and merits at most a trivial fine.
Harry First: Sure to both. Note that although violations of Article 82 of the EC Treaty are not technically criminal, Microsoft was fined a substantial amount—the estimate is that the fines will eventually exceed $2 billion—and the CFI upheld a large part of those fines (more than $600 million) in its decision.
Is this a modern day example of inquisition logic - impeccable argument drafted by esteemed but technically ignorant authorities who are simply convinced that witches can fly? During the entire discussion, not once has any one been able to sustain a need to have access to intellectual property data not already available in the
innards of the existent documentation for WIN32 APIs. Are perhaps those who clamor for details simply lazy or worse?
Ronald Barker, Munich, Germany
Richard Epstein: I have no idea what a WIN32 API is, but I assume that it has something to do with the inner operation of the Microsoft system. I don’t see the case for what the commission did, because I think, as mentioned in my answer to Neil that the US settlement was about right. The key insight is that dominant positions are par for the course in network industries. The American strategy called for competition above the Operating System. The EU system calls for the subsidy of Microsoft competitors to create a second operating system. But even if two (or more) came into view, dominance would reassert itself, as both users and application writers like to have a single platform on which they can transact business.
Harry First: I can’t speak for the logic of the inquisition, nor for the technical abilities of the European Commission’s staff or the CFI, nor even for the documentation for WIN32 APIs. What I can say is that the technical experts both in the US and in Europe who have looked closely at what Microsoft has been required to produce have concluded that producing this documentation is not an easy task, but also the Microsoft’s efforts to have been, shall we say, less than useful. I’m not sure what that makes Microsoft.
About the experts:
Professor Harry First has been a faculty member at New York University School of Law since 1976, teaching courses in antitrust, regulated industries, business crime, and innovation policy. He is currently working on a book titled Microsoft and the Globalization of Competition Policy: A Study in Antitrust Institutions. From 1999 to 2001 Professor First served full-time as Chief of the Antitrust Bureau in the Office of the New York State Attorney General.
Richard A. Epstein is a Professor of Law at the University of Chicago, and a Senior Fellow at the Hoover Institution. He has extensive ties with Microsoft and with their financial support has written Antitrust Consent Decrees in Theory and Practice: Why More is Less, forthcoming from the American Enterprise Institute.
Copyright The Financial Times Limited 2016. You may share using our article tools.
Please don't cut articles from FT.com and redistribute by email or post to the web.
Sign up for email briefings to stay up to date on topics you are interested in