The Selection Hypothesis and the Relationship between Trial and Plaintiff Victory

The selection hypothesis of Priest and Klein explains the selection of cases for trial, from the underlying population of filed cases, based on the position of the legal standard, the degree of stake asymmetry, and the predictability of trial outcomes. This paper develops implications of the selection hypothesis for the relationship between trial rates and plaintiff win rates. We find strong evidence for the selection hypothesis in estimated relationships between trial rates and plaintiff win rates at trial across case types and judges. We then structurally estimate the model on judge data, yielding estimates of the model's major parameters (the decision standard, the degree of stake asymmetry, and the uncertainty parameter) for each of three major case types, contracts, property rights, and torts. We are able to infer that tried cases are unrepresentative of filed cases and that stakes are higher for plaintiffs in contract and property rights cases and higher for defendants in tort cases. Finally, we infer that the uncertainty surrounding case outcomes is higher for tort cases than for the others, supporting the view of tort system critics that legal standards in tort cases are not clearly understood.