Educational Commonality: A new Metric for Measuring the Intensity of Relationships in the Global Financial Network
This project aimed to develop a novel metric for measuring the intensity of relationships between actors within the global financial network, with the ultimate goal of shedding light on systemic financial exploitation. Using Item Response Theory (IRT), specifically focusing on Educational Commonality (ECOM) as a latent variable, the project successfully constructed a unique measurement model. The study focused on FDIC-insured banks, utilizing data related to housing investments.
The initial assumption that a measurement model could be built without actual data proved impractical, leading to the manual creation of a dataset comprising information on board members of 262 FDIC-insured banks. The dataset included details on educational backgrounds, such as undergraduate and advanced business degrees, and law degrees, which were used to measure ECOM.
The modeling process involved using JAGS to estimate coefficients for ECOM based on the rank of undergraduate education, advanced business degree, and law degree for each bank. Despite the unexpected complexity, the generated values aligned with expectations, demonstrating that banks with higher ECOM values tended to have more board members with degrees from top schools.
While the project faced challenges, including a reduced sample size and the need for data collection, it ultimately proved effective. The extension granted was attributed to the unforeseen intricacies of the modeling process, requiring a deeper understanding of Item Response Theory.