In a very brief (7 pp.) technical--though informative--paper circulating on SSRN, Empirical Studies of Copyright Litigation: Nature of Suit Coding, Matthew Sag (Loyola-Chicago) assesses variable coding reliability in a commonly-used database for legal scholars (PACER). Specifically, the paper focuses on "Nature of Suit" variable in the PACER records for empirical studies of copyright litigation. While Sag finds that the variable does not, in fact, capture all copyright cases, it nonetheless remains sufficient "for most purposes." Sag notes that the variable is especially suspect for copyright cases that involved pro se litigants and where copyright was not the primary litigated issue. Sag estimates that the "820 code" captures "80 to 85% of true copyright cases leading to written opinions."