Document Type

Article

Publication Title

Florida State University Law Review

Abstract

Common wisdom holds that many federal judges do not write their own opinions. While their degree of input into opinion writing varies, almost all rely to some extent on law clerks, typically recent law school graduates, to research and draft substantial sections of the opinion. Why should we care which judges write their opinions? We posit that determining the actual input of federal judges into the authorship of opinions provides useful information in a number of contexts, including judicial promotion decisions, the allocation of scarce judicial resources, and the judicial clerkship market for law students. We start with generic tests of authorship obtained from computational linguistics. To assess the effectiveness of these tests in the judicial context, we identify (based on an informal survey of several sitting federal judges), three judges reputed to write their own opinions. We then use a randomly selected set of opinions for active circuit judges from 1998 to 2000 to examine whether the generic tests succeed in ranking these three judges high in terms of authorship. The generic tests failed to distinguish authorship due in part to the subject matter specific nature of opinions. Whether an opinion is a criminal, administrative, or commercial law opinion matters. Comparing two judges based on their criminal law opinions may provide different results than comparing the same judges based on the criminal law opinions of one judge against the administrative law opinions of the other judge. Our generic tests did not control for subject matter of opinions. We provide a number of more customized tests for the judicial context that attempt to control for subject matter. Using these customized tests, we are able to distinguish the three test judges based on authorship.

First Page

1077

Volume

32

Publication Date

2005

Download

Share

COinS