Define one's own competences

In the field of research of law and courts, great effort is made to understand how constitutional courts determine their relationship with the legislator. To this end, innumerous court decisions are manually coded based on their outcome. This research paper aims to follow up on the aforementioned by exploring the influence of the German Federal Constitutional Court in particular. However, the focus will not be on the outcome of the single decision, but on the specific reasoning based on which the court grants legal competences in its case law. In order to obtain this information, an automated approach will be used that includes linguistic features such as part-of-speech tagging and morphology. To assess how the Federal Constitutional Court defines competences requires considering two dimensions: On the one hand, the court can define the competences of the legislator. In doing so, it has several options: The Constitutional Court can prohibit or oblige the legislator, but it can also indicate room for discretion. On the other hand, the court can, at the same time, define its own competences in that it can decide for itself what it is allowed to control. For example, in a well-known ruling on basic welfare benefits, the German Constitutional Court states that the legislator has great leeway for making decisions in the field of social law, but that the court still has the authority to control these same decisions. The paper wants to put these two dimensions into a formal model and automatically measure how the court uses them. 3400 decisions are automatically studied based on rules created here for that rely on linguistic information. This rule set is used to detect and classify which actor is allowed to do what. The large language models of computational linguistics will thus be made fruitful for the analysis of jurisprudence in particular as well as law and courts research in general.