Fraud in US state and local government programs (unemployment insurance, Medicaid, SNAP, tax refunds, loans, grants, and identity theft) persists largely because relevant data is fragmented across siloed departmental systems that weren't designed to share information. Consolidating and normalizing existing data, rather than replacing systems, allows investigators to spot cross-program patterns like shared bank accounts or mismatched addresses that would otherwise go undetected. California's EDD is cited as an example, having used Elastic to unify data across 3,000+ servers, achieving a 99% reduction in mean time to response and securing over 850 billion records. Practical starting points include data-sharing agreements, platform audits, and entity resolution capabilities, with incremental rollout recommended over large-scale overhauls.
