Professors Kagan, Gill and Marouf have identified a remarkable gap in the Westlaw and Lexis databases. While those databases include all decisions designated as “published” and some other less elaborate, less detailed, decisions designated as “unpublished,” many decisions are not included at all.
The authors discovered this by studying immigration decisions on PACER. The cases involved review by the U.S. Courts of Appeals of administrative decisions by the Board of Immigration Appeals in the Department of Justice. In some cases, no appellate rulings were available, because they were sealed, for example, or because the case was resolved by a short docket entry. But even where merits decisions were issued and publicly available, many did not appear in the searchable databases, although Lexis had far more than Westlaw. Of course, in such cases the decisions are available on PACER and can be found by docket number on the Lexis and Westlaw mirrors of PACER. But the whole point of a searchable database is to find cases one does not already know about.
As good scholarship often does, this discovery raises a number of questions. The first is whether there are other categories of cases that are also selectively reported. I regularly teach a federal criminal appellate clinic and I have been surprised that dispositions that cite cases and vacate judgments do not find their way into the databases.
A second question is why Lexis and Westlaw would choose to deprive researchers of useful information. This is a particular problem because one party, the Department of Justice, has ready access to all the decisions in their own files. Accordingly, without complete coverage, attorneys representing immigrants (or criminal defendants) will be able to see only a partial picture of what the courts are doing.
The paper nicely explains the constitutional controversy surrounding published and unpublished opinions, in particular whether it is permissible for an Article III appellate court to issue non-precedential decisions. But even if the invisible adjudications are legitimately non-precedential, they remain significant. Lawyers want to know what the judges they appear before have actually done in similar cases; even if those judges are not required to apply the same law in the future, the odds are that they will. Similarly, scholars trying to understand the courts’ views on particular legal problems will often find a non-binding opinion just as illuminating as a precedential one.
Unlike many problems addressed by scholars, this one is easy to fix. Lexis and Westlaw should be non-selective in incorporating dispositions by the U.S. Courts of Appeals into their databases. An affirmance, reversal, vacation, or remand, however designated, whether stipulated, procedural, on the merits, or otherwise, should be in the database. Odd, quirky dispositions will be useful for lawyers in the future with odd, quirky cases. Existing search functions allow limiting of results to reported cases; perhaps it could be tweaked further to account for this new category of decision.