The global financial crisis which began in 2007 is the most severe economic event since the 1930s. The profound political and economic consequences of the crisis have clarified the need to better understand the financial system at both micro and macro levels. This dissertation advances research on both fronts. First, it utilizes network prominence measures to look at the pre- and post-crisis organization of the global banking system, finding that American prestige has increased as a result of the crisis. Second, it employs complex network theory and inferential statistical models to explain why the global banking system is organized as it is, finding that endogenous processes interact with monadic and dyadic political economy variables to produce a global structure. Third, it examines bank behaviors at the firm level, demonstrates that representative agent models are insufficient for explaining the patterns observed, proposes an alternative approach drawing from ecological finance theory, tests the model using Bayesian regression, and finds support for the new approach. In sum, this dissertation demonstrates the need for further quantitative political economy work at both the micro and macro levels of the global financial system and provides several possible pathways forward.