Ahad Ahadi has posted When Laws Contradict Themselves: A Quantitative Audit of U.S. Tax Code Coherence on SSRN. Here is the abstract:
Problem. Complex legal systems accumulate internal contradictions that undermine fair enforcement and consistent compliance. Why it matters. We currently lack quantitative methods to measure where and how severely statutory authorities conflict-making prioritization of legislative fixes nearly impossible. What I built. I developed CALE-LAW, a neuro-symbolic framework that computes contradiction probability (C), ambiguity entropy (H), and system-level coherence (κ) across legal authorities, with impact scoring (Δκ) to identify high-leverage clarifications. Key finding. Applied to simulated U.S. tax code sections, the system identified authority pairs with C>0.70 and H>0.95, demonstrating that fundamental definitional conflicts (e.g., income thresholds for different programs) create enforcement impossibilities-not mere complexity.
Very interesting.