Abstract
What are the positions of highest courts towards political reforms? In order to answer this question scholarship on the US Supreme Court uses judicial votes but those are often not available in cross-country comparison. We present an approach scaling judicial decisions based on textual features other than votes. In particular, we argue that the legal outcome and tendencies field in briefs by scalable political actors can be used to compute a vote-matrix similar to roll-call data in legislative studies. With the obtained data we are able to estimate location scores employing a standard two-parameter item-response theory model using R STAN. Moreover, computing priors from the known manifesto positions of political actors allows as to meaningful anchor the court’s locations in relation to the political positions in one common policy-space. The procedure applies independently of specific judicial systems but to outline the feasibility, we assess 100 senate decisions by the German Federal Constitutional Court. Our computed spatial measures allow for assessing the judiciary’s willingness to align or diverge from the opinion of governing parties and their political reform. Moreover, the measure helps to validate common theoretical expectations which are based on the assumption that courts are strategic (political) actors.

To what extent is a constitutional court an integral actor in the system of checks and balances? In order to answer this question, we present the novel Constitutional Court Database (CCDB) which links (1) 2006 Senate decisions (2) combining 3284 different proceedings referred to the German Federal Constitutional Court (GFCC) between 1972 to 2010 to (3) information from the political environment and (4) to societal developments. The relational structure of the database is well suited to connect information across the four different layers in flexible and dynamic ways. This allows to take different perspectives on the GFCC as a legal, political or societal actor and as a representative case of a highest court with constitutional review powers.

In order to outline the usability of the database, we present a “how-to” paper. Designing the relational, multi-layered database a number of theoretical, conceptual and computational challenges had to be overcome. Presenting those challenges and details on how to compute the database will help scholars of political science and methodologists alike. Moreover, connecting the judicial and legislative domain to address the question to what extent legislative length is a predictor of a law’s referral to a constitutional court, we present a use-case to outline the power of the database. The flexible CCDB extents on existing static databases, such as the Supreme Court database. The aim of the novel data-structure is to provide a tool useful to legal scholar, scholars of political science, journalists and the public.

Status
Draft presented.

Together with
Thomas Gschwend (University of Mannheim)
Christoph Hönnige (Leibniz University Hannover)
Caroline E. Wittig