The thirteen circuits' thirteen different multifactor tests for the likelihood of consumer confusion have long played a role of central importance in American trademark litigation, yet they have received little academic attention and no empirical analysis. Courts, commentators, and practitioners have all the while speculated about which factors, if any, drive the outcome of the tests, how the factors interact, and, most importantly, whether the circuits' different tests, given the same facts,

more »

... uld yield different outcomes. With a view to the settling of these questions and ultimately to the reform of the multifactor tests, this Article sets forth the results of an empirical study of all reported federal district court opinions for the five-year period from 2000 to 2004 in which a multifactor test for the likelihood of consumer confusion was used. In the process, it presents the multifactor test for the likelihood of consumer confusion as an ideal case study in legal multifactor decision making and develops a methodology and theoretical toolkit for the study of this form of legal analysis across the many areas of law that employ multifactor tests. Working from an original data set of 331 opinions, this Article finds significant variation among the circuits in the application and outcome of their respective tests. Drawing upon recent social science learning on cognition and decision making, it further shows that judges employ "fast and frugal" heuristics to short-circuit the multifactor analysis. A few factors prove to be decisive; the rest are at best redundant and at worst irrelevant. Judges tend to "stampede" these remaining factors to conform to the test outcome, particularly when they find infringement. This Article also makes a variety of factor-specific findings that contravene conventional wisdom in trademark law. For example, courts typically declare that no single factor outcome is dispositive. The data clearly contradict this assertion. A finding that the similarity of the marks factor does not favor a likelihood of confusion is, in practice, dispositive, and a finding that the proximity of the goods factor does not favor a likelihood of confusion is nearly dispositive. Relatedly, at least one circuit, the Second, has declared the defendant's intent factor to be "irrelevant." Meanwhile, the data show that a finding of bad faith intent is in fact dispositive, as much in the district courts of the Second Circuit as in those of any other. Furthermore, many believe survey evidence to be the best and most persuasive form of evidence of the likelihood of confusion. The data reveal, however, that surveys are rarely presented by parties or credited by courts. Additionally, this Article finds that the venerable doctrine of trademark strength-and in particular of inherent distinctiveness-has broken down. The data show a remarkable divergence between traditional circuit court (and casebook) doctrine on this and current district court practice.