Most U.S. states have enacted laws authorizing or requiring financial institutions to report and/or prevent suspected elder financial exploitation (EFE). This paper develops a structural model of the interaction between reports of suspected EFE and actual incidences of EFE to evaluate the effectiveness of these laws. Because state EFE laws vary widely in scope, application, and the financial institutionssubject to them, I use an Mnet penalized regression - a supervised machine learning method for variable selection - within a difference-in-differences framework to identify which legal provisions drive changes in reporting and prevention outcomes.
with Arvind Mahajan and Rajendra Patidar
with Boone Bowles