Michael Findley (UT-Austin), Daniel Nielson (UT-Austin) and Jason Sharman (Cambridge)
Banking Bad: A Global Field Experiment Testing Money-Laundering Risks in International Finance
Abstract
This article bridges the disciplinary divide between behavioralism and institutionalism via a global field experiment assessing the effects of organizational scripts on banks’ and intermediaries’ reactions to requests for anonymous access to the financial system. We thus test the effectiveness of global banking regulations designed to screen out criminal money based on risk assessment. Directly contrary to both the international regulatory regime and scholars’ surveyed expectations, banks prove remarkably insensitive to risk in opening accounts. We attribute this insensitivity to the power of organizational scripts: shared, standardized patterns of institutionally constrained agent behavior for diagnosis and response. Smaller, less regulated intermediary firms appear less influenced by scripts and more sensitive to varying customer risk. Because this study’s experimental conditions and subject pool operate at the level of institutions (firms), rather than individuals, it mitigates the disciplinary divide between experimental methods testing behavior and theories of institutions. As a global field experiment, the study also addresses common external-validity limitations.