categorize the lawsuits according to the scheme we used for coding cases in our database. Figure B.2 compares the FJC data and the data we collected for this study.
There are a number of explanations for the differences between the FJC study data and our own. First, differences exist in the scope of the two studies: The FJC study was limited to four federal district courts, whereas our study reflects the nationwide population of class actions, including class actions brought in state courts. Second, the FJC study identified and described class action law- suits from court records. Our study relies on less inclusive appellate decisions and more inclusive but less precise newspaper reports. Third, the FJC study identified cases terminated in 1992-1994, some of which were filed years earlier. Our study captures class actions in process during 1995-1996, which include newly filed and terminated lawsuits as well as cases still pending.