Building on prior work that quantified the extent to which the responsibilities of regulating financial markets, institutions, and activities in the EU have been delegated to bureaucratic agencies from 1964 through 2017 (O’Halloran et al., 2018), the Capstone team this year focused on identifying specific aspects of the legislative process in the EU that have produced the observed delegation and discretion outcomes. Both phases of work lay the foundation for the construction of a classification and regression tree (CART) algorithm that can be used to build a predictive model for delegation and discretion in the EU.

The Capstone team first assessed the indices of delegation and discretion levels for EU financial regulations created by O’Halloran et al. (2018) using techniques in natural language processing (NLP) and diagnostic analysis. Through process-tracing for regulations and directives that entered into force from 2004 through 2017, the Capstone team then explored potential causal mechanisms in the EU legislative process that produced the observed delegation and discretion outcomes.

Contrary to the concerns regarding transparency and representation in EU-level policy-making processes, the Capstone team’s analysis of the process-tracing data reveals that the legislative process in the EU is quite porous and receives inputs from many sub-level actors. At the same time, the dispersion and inconsistent availability of information, as well as the labyrinthine organization of documentation of policy-making processes only reaffirm the need to increase the transparency of EU institutions to citizens.