Because my Cornell colleagues, Ted Eisenberg, Valerie Hans, and Marty Wells, are all too modest to shamelessly self-promote, the task falls to me to do so on their behalf (with neither their consent nor knowledge). Their recent paper, The Relation Between Punitive and Compensatory Awards: Combining Extreme Data with the Mass of Awards, is now available on SSRN. As the excerpted abstract, below, suggests, this paper intrigues for methodological and substantive reasons. On the methods front, the paper cleverly exploits three separate data sets by integrating them into one. This move facilitates studying the compensatory:punitive damage awards relation throughout a broad range of trial awards. Substantively, what they find, in part, is that even after accounting for "extreme" punitive awards, compensatory awards still explain the bulk of variation in punitive awards.
"This article assesses the relation between punitive and compensatory damages by combining two data sets of extreme awards with state court data from the National Center for State Courts (NCSC) for 1992, 1996, and 2001. One data set of extreme awards consists of punitive damages awards in excess of $100 million from 1985 through 2003, gathered by Hersch and Viscusi (H-V); the other includes the National Law Journal's (NLJ) annual reports of the 100 largest trial verdicts from 2001 to 2004.... Combining the data sets assists in observing the punitive-compensatory relation throughout the full range of trial awards.... The models indicate that the compensatory award explains more than 50 percent of the variance in the punitive award. We also find no increase in punitive or compensatory awards over time in any of the three data sets."