Robert G. Bone

Abstract

This Article addresses the normative issues raised by the use of statistical sampling to adjudicate large case aggregations. In Wal-Mart Stores, Inc. v. Dukes, the Supreme Court referred to sampling pejoratively as “Trial by Formula,” but this Article argues that the label is undeserved. Sampling can be justified in many more situations than courts currently recognize, and society is paying a high price for limiting its use. Expanding on my earlier work, Statistical Adjudication: Rights, Justice, and Utility in a World of Process Scarcity, this Article develops the analysis in four ways. First, it examines the effect of sampling on settlement and explores how sampling influences frivolous and weak filings; although sampling may reduce settlement likelihood and provide cover for undesirable lawsuits, these effects do not outweigh its benefits in sufficiently large aggregations. Second, it evaluates sampling under an outcome-oriented, rights-based theory, noting that the most serious issue is that sampling gives high-value plaintiffs only an average recovery; the Article generalizes this concern beyond the earlier treatment. Third, it offers additional reflections on process-based participation and the day-in-court right based on more recent scholarship. Fourth, it explores a further objection not addressed previously—the “methodological legitimacy objection”—which argues that adjudication fundamentally requires reasoned, fact-sensitive deliberation, whereas sampling substitutes formulaic methods for individualized reasoning. Although this objection has intuitive force, the Article demonstrates that it is difficult to defend rigorously.

---