No RSVP required for current Princeton faculty, staff, and students. Open to members of the public by invitation only. Please contact Jean Butcher at if you are interested in attending a particular lunch.
“Probable cause” is not about probability. It is about plausibility. To determine if an officer has the requisite suspicion to perform a search or seizure, what matters is not the statistical likelihood that a “person, house, paper or effect” is linked to criminal activity. What matters is whether criminal activity provides a convincing explanation of observed facts. For an inference to qualify as plausible, an observer must understand why the inference follows; she must be able to explain its relationship to the facts. Probable inferences, by contrast, do not require explanations. An inference can be probable—in a predictive sense, based on past trends—without a human observer understanding what makes it so.
In many cases, plausibility and probability overlap. An inference that accounts for observed facts is often likely to be true, and vice versa. But there is an important sub-set of cases in which the two properties pull apart, raising deep questions about the underpinnings of Fourth Amendment suspicion: inferences generated by predictive algorithms. In this Article, I argue that casting suspicion in terms of plausibility, rather than probability, is both more consistent with established law and crucial to the Fourth Amendment’s normative integrity. Before law enforcement officials may intrude on private life, they must explain why they believe wrongdoing has occurred. This “explanation-giving” requirement has two key virtues. First, it facilitates governance; we cannot effectively regulate what we do not understand. Second, it allows judges to consider the “other side of the story”—the innocent version of events a suspect might offer on her own behalf—before warranting searches and seizures. In closing, I connect these virtues to broader themes of democratic theory. In a free society, legitimacy is not measured solely by outcomes. The exercise of state power must be explained—and the explanations must be responsive both to the democratic community writ large and to the specific individuals whose interests are infringed.
Kiel Brennan-Marquez is a postdoctoral research fellow at the Information Law Institute at New York University and a Visiting Fellow at the Information Society Project at Yale Law School. His main research interest is how technological evolution, especially related to big data, is reshaping the judicial process. Recently, his scholarship has focused on the rise of “data-driven policing,” and the way changes in the collection and use of information by law enforcement force us to rethink the normative foundations of procedure and evidence law. Kiel holds a B.A. in philosophy and religious studies from Pomona College, and a J.D. from Yale Law School. He clerked for the Honorable Shira A. Scheindlin of the Southern District of New York, who presided over Floyd v. City of New York, the widely-publicized challenge to the NYPD’s (now reformed) stop-and-frisk program.