The Promise and Pitfalls of Automated Text-Scaling Techniques for the Analysis of Judicial Opinions

Scaling Court Decisions with Citation Networks

“…Existing approaches pay close attention to studying the language of court decisions and have been exploring supervised and unsupervised text-scaling (e.g. Dyevre, 2019;Evans et al, 2007;Jakab, Dyevre and Itzcovich, 2017;McGuire and Vanberg, 2005) or dictionary methods (e.g. Wedeking, 2011, 2012).…”

Delegation at the U.S. Federal Appellate Courts: The Power to Remand As a Double-Edged Sword

“…Our first part of the textual analysis uses the algorithm "Wordscores" (Laver et al 2003), originally designed to analyze the views of politicians but recently adapted to many other contexts, including judicial decisions (see Dyevre 2015, and various papers mentioned therein). In a nutshell, the algorithm relies on two (or more) "virgin" master files, each corresponding to a different point (usually the edges) along the dimension of interest.…”

Identification, Categorisation and Forecasting of Court Decisions

“…For example, Hunter et al (2008, 79) state: "This tradition has not been established in the United Kingdom, perhaps because we do not have a sufficient number of judges at the appropriate level who are not male and white to make such statistical analysis worthwhile". Still, researchers have applied quantitative methods to datasets of case law from, for example, Belgium (De Jaeger, 2017), the Czech Republic (Bricker, 2017), France (Sulea et al, 2017), Germany (Dyevre, 2015;Bricker, 2017;Hartung, 2021), Israel (Doron et al, 2015), Japan (Kyo, 2022), Latvia (Bricker, 2017) (Bricker, 2017), Portugal (Rodrigues and Campina, 2021), Slovenia (Bricker, 2017), Spain (Garoupa et al, 2012), and Sweden (Derlén and Lindholm, 2017c).…”

“…Nevertheless, a number of researchers are now using computerised techniques to collect case law and automatically generate usable information from it (see Trompper Ruggeri et al, 2022). For example, Dyevre (2015) discusses the use of automated content analysis techniques in the legal discipline, using tools such as Wordscores 4 and Wordfish, 5 which have traditionally been used to automatically extract political positions, by using word frequencies in text documents. The author applied the two techniques to the analysis of a (relatively small) dataset of 16 judgements on European integration by the German Federal Constitutional Court.…”