By Jason Rantanen
Many empirical studies of Federal Circuit jurisprudence rely on searches of one of the leading legal databases such as Westlaw or Lexis.1 Relying on a search of a single database is potentially problematic, however, if the substantive content of the databases is not identical – in other words, if Lexis and Westlaw don't contain the same universe of cases, any claims about the results are necessarily limited by the dataset being used.
In order to avoid this problem in connection with an empirical study of inequitable conduct that I'm working on, I recently performed a comparison of results obtained by a keyword search on Westlaw versus an identical search on Lexis. This comparison revealed that although the data obtained from the two sources is largely comparable, small differences do exist that certain types of searchers may want to take into account.
The chart below shows the results, by year, produced by searching for the term "inequitable conduct" in the Federal Circuit databases for Lexis and Westlaw.2,3 In large part, the results suggest that there is relatively little variation between the two databases: in the context of this analysis, only 30 out of 681 (4.4%) Westlaw opinions were not found in Lexis and only 6 out of 657 (0.9%) Lexis opinions were not found in Westlaw. Furthermore, in no instance did a precedential opinion appear in one database but not another. Rather, all of the single-database hits were nonprecedential opinions. In addition, nearly all of the nonprecedential opinions appearing in only a single database issued prior to 1991. That said, there is a significant difference with respect to nonprecedential opinions for the period prior to 1991: nearly all the nonprecedential opinions from this time period are found in only one database.4 Thus, while it may not be efficient for most practitioners to run keyword searches on multiple databases, scholars conducting empirical research on Federal Circuit decisions, or individuals searching for older party-specific nonprecedential opinions, may want to be sensitive to the differences between the two databases.
Figure 1: Hits Resulting from Search Term "Inequitable Conduct" in Lexis and Westlaw5
1 See, e.g., Christopher A. Cotropia, "Determining Uniformity Within the Federal Circuit By Measuring Dissent and En Banc Review," 43 LOY. L.A. L. REV. 801, 811 (2010); Christian E. Mammen, Controlling the "Plague": Reforming the Doctrine of Inequitable Conduct,24 BERKELEY TECH. L.J. 1329, 1348 (2009); Craig Allen Nard, Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit Patent Law Jurisprudence, 39 HOUS. L. REV. 667, 678 fn 40 (2002); Christian A. Chu, Empirical Analysis of The Federal Circuit's Claim Construction Trends, 16 BERK. TECH. L.J. 1075, 1092 (2001); but see R. Polk Wagner and Lee Petherbridge, Is the Federal Circuit Succeeding? An Empirical Assessment of Judicial Performance, 152 U. PA. L. REV. 1105, 1145 (2004) (using results from both Lexis and Westlaw). In addition, some early empirical studies used the United States Patent Quarterly as their source (see, e.g., John R. Allison & Mark A. Lemley, How Federal Circuit Judges Vote in Patent Validity Cases, 10 FED. CIR. B.J. 435, 436 (2001)), which apparently contains some, but not all, nonprecedential patent decisions. See e.g., Wirco, Inc. v. Rolock, Inc., 17 U.S.P.Q.2d 1084, 1990 WL 12901 (Fed. Cir. Feb. 14, 1990) as opposed to Mechanical Plastics Corp. v. Unifast Industries, Inc., 1988 WL 25422 (Fed. Cir. March 28, 1988).
2 For purposes of creating this chart, I used the additional limitation "CO(c.a.fed.)" to exclude CCPA and Court of Claims decisions, some of which are included in Westlaw's "ctaf" database. In addition, I separately compared the results from a keyword search of the entire "ctaf" database to the results from the Lexis search in order to identify any Lexis-only hits that might have been excluded by the Westlaw court limiter. This process did not reveal any such hits.
3 These results include every search hit that contained the term "inequitable conduct," regardless of the context. It thus should not be viewed as reflecting any measure of substantive inequitable conduct determinations. In addition, hits resulting from search terms appearing only in the database-specific casenotes were excluded, as were instances where the database contained duplicate entries of the case opinion or included a copy of a withdrawn opinion. For more substantive analyses of inequitable conduct jurisprudence, see Lee Petherbridge, Jason Rantanen & Ali Mojibi, The Federal Circuit and Inequitable Conduct: An Empirical Assessment, ssrn id 1686102; Mammen, supra note 1.
4 In addition to double-checking the single-database cases against the search results of the other database, I also spot-checked several decisions against the Lexis/Westlaw case database as a whole by using terms from the parties' names and the Federal Reporter citation, where available. Although in some instances this resulted in a hit, the contents of those hits were blank. See, e.g., Air Products & Chemicals, Inc. v. Tanner, 738 F.2d 454 (1984) (nonprecedential) (no opinion in Westlaw; opinion in Lexis).
4 Although the results for 1997 appear equal, each database contained one opinion for that year that is not in the other database.