THE DETERMINATION OF ANTITRUST LIABILITY IN UNITED STATES v. MICROSOFT: THE EMPIRICAL EVIDENCE THE DEPARTMENT OF JUSTICE USED TO PROVE ITS CASE
Correspondence: * Research Economist for the U.S. Department of Justice. Email: wayne.dunham{at}usdoj.gov. The author was assigned to the Microsoft investigation and case from mid-1996 through the entry of the final judgment in November 2002. This paper represents the views of its author and does not represent the views of the U.S. Department of Justice. All usual disclaimers apply. The author thanks Timothy Bresnahan, Franklin Fisher, Rick Warren-Boulton, Dan Rubinfeld, Ken Heyer, Russ Pittman, and others at the Department of Justice who reviewed this paper for style, content and accuracy as well as the Department of Justice's Microsoft trial and appellate teams, including Diane Owen, Karma Giulianelli, Phillip Malone, David Seidman, David Sibley, Greg Werden, and the numerous others with whom he has discussed the content of the paper.
This paper considers the empirical evidence used by the Department of Justice in the U.S. v. Microsoft antitrust case to prove that Microsoft engaged in exclusionary (and anticompetitive) actions in the browser market as part of its efforts to maintain its dominance of the personal computer operating system market. This evidence deserves special consideration because the District Court made the unusual decision to rely on the empirical evidence presented by the Department of Justice rather than the empirical evidence presented by Microsoft. This decision was unusual because Microsoft's evidence had a strong presumption of validity as it was based on data that Microsoft collected and used in the ordinary course of its business. Furthermore, no market participants used the Department of Justice-sponsored data in any meaningful way. Although it is impossible to determine with any certainty why the District Court ruled the way it did, I conclude that there were two driving forces in the court's decision. The Department of Justice identified serious flaws in Microsoft's data, making it unreliable for the purposes for which Microsoft was using it in the trial. The Department of Justice was also able to show that no such flaws affected the data it sponsored and indeed, on many points, that data was more consistent with the testimony of Microsoft executives than the data sponsored by Microsoft.