Saliency, Anchors & Frames: A Multicomponent Damages Experiment

Modern technology products contain thousands, sometimes hundreds of thousands, of different features. Nonetheless, when electronics manufacturers are sued for patent infringement, these suits typically accuse only one feature, or in more complex suits, a handful of features, of actual patent infringement. But damages verdicts often do not reflect the relatively small contribution an individual patent makes to an infringing product. One study observed that verdicts in these types of cases average 9.98% of the price of the entire product. While both courts and commentators have blamed the law of patent damages, the role cognitive biases play in these outsized damages awards has been understudied. Relying on decision-making concepts from other contexts, we hypothesize that two biases, namely, a saliency bias and anchoring, may be at work in a patent trial. Since the infringing feature is the most salient feature in a patent trial (i.e. the focus of the trial), jurors may tend to overvalue that feature. Moreover, a patentee’s irrationally high damages demand may “anchor” the juries to that number. We conducted an online 3x3x2 between-subjects experiment to test whether these biases exist and if so, whether particular debiasing techniques may reduce these biases. In eighteen different scenarios, mock jurors were asked to assess damages for different smartphone features. The three manipulations involved: 1) rotating three features so that they were either the feature underlying the plaintiffs’ claim (the “feature-in-suit”) or one of the other features defendant identified as contributing to the smartphone’s overall value; 2) changing the jury verdict form so that mock jurors had to evaluate both the feature-in-suit and other features together; and 3) having the defendant explicitly call out the plaintiff for anchoring the jury in an irrationally high number. The results suggest that some combination of the saliency bias and anchoring were at play when juries assessed damages for all three tested features. However, for the storage feature the results were only significant for the feature’s relative rank, but not its dollar valuation. That may be because mock jurors are familiar with the cost of the increased storage. Modifying the jury verdict form reduced, but did not eliminate, the primary effect of the saliency bias, while the defendant’s tactic of exposing the plaintiffs’ anchor did not significantly reduce damages. In addition, qualitative comments suggested that some mock jurors resisted the jury instructions designed to compensate plaintiffs for the missing feature and instead assessed damages to punish the defendant.