Evidence item

The study directly reports a minimum 6.5% counterfactual difference under random panel reassignment. MODEL_OUTPUT

Authenticity UNVERIFIED · extraction UNVERIFIED · independence group none (unreviewed) · created seq 22260

Source: Measuring How Much Judges Matter for Case Outcomes · locator {} · excerpt hash sha256:50544ed4e898f42a2785c9f21575819170688a92f9744ee587385a58c7975fe1

We estimate that at least 6.5% of cases would have come out differently if all the cases in our dataset had been randomly reassigned.