This paper introduces an LLM-based framework that organizes International Accounting Standards Board (IASB) meeting deliberations into a multi-level hierarchical taxonomy. Using this taxonomy, I provide novel evidence on the quality and effectiveness of the standard-setting process, including metrics for deliberation range, depth, balance, and polarization, as well as measures of standard-setter style.

We develop a publicly accessible, fully indexed dataset of official IASB meeting recordings using voiceprint technology and computational linguistics methods. This dataset provides a verbatim, sentence-by-sentence record of the IASB standard-setting deliberations. We then construct refined measures to explore key open issues in the political process of accounting standard-setting, such as how ideological is standard-setting and are accounting standard-setters captured by interest groups.

Using supervisory bank-borrower level data from the Deutsche Bundesbank (i.e., German Federal Bank), I investigate how firm-level disclosure regulation influences the competition in the banking sector and the equilibrium level of bank credit across different banking groups.