Do Lawyers Know Uncertainty when They See It?

Abstract: Lawyers are supposed to be experts at assessing legal risk. We analyze how well they can predict the decision of a court on a legal issue. In our model, legal uncertainty arises from random errors of judges and actors (or their advisers) in determining the applicable legal standard. This implies that lawyers will often produce widely dispersed probability estimates regarding the outcome of a case. Our theoretical model yields a number of testable implications. One of them is that probability estimates are strongly related to a lawyer’s own judgment of the case. In contrast to much of the previous literature, this “consensus effect” does not result from a cognitive bias but reflects limited knowledge regarding the applicable legal standard. An experiment with four different cases and 215 law students largely confirms the predictions of the model.