Making Assets Contractible: Artificial Intelligence, Identity Verification, and Transaction Costs in Livestock Insurance

Date:  Tuesday, 18 August 2026
Time:  3:30 PM – 5 PM
Venue:  Meeting Room 5 (ABS-02-015) 


Chairperson: Assoc Prof Jinggong Zhang


Abstract
Many valuable assets, such as livestock, are difficult to insure or finance because their identity cannot be reliably verified at low cost. AI-based biometric technologies can relax this constraint by making asset verification more accurate and less costly. This paper studies how such technology affects identity verification and transaction costs in the context of livestock insurance. We first develop a theoretical framework in which weak identity verification allows farmers to claim losses on uninsured cattle, so that a small number of insurance contracts can cover losses arising from deaths across a larger herd. This distortion generates a downward-sloping marginal-cost curve that mimics adverse selection even when animals are homogeneous, leading to inefficient under-enrollment. Costly physical verification under current ear-tag technology also constrains underwriting capacity and limits the scale of insurance provision. Using administrative data from 2021–2025 covering more than two million cow-years, approximately 48,000 policyholders, and 1,700 underwriters, we find that AI exposure reduces claim rates by 0.77 percentage points (14 percent), and increases insured herd size by 0.81 cows (5 percent). We also show that AI sharply reduces underwriting time per cow. By reducing the time and physical effort required for identification, AI enables underwriters to serve more policyholders, insure more cattle, and gradually extend coverage to more distant farmers. Together, these findings show that AI can reduce missing markets by making hard-to-verify assets contractible at scale.  Back-of-the-envelope calculations imply that, over the first three years of AI adoption, the technology reduced claims by about 7,750 and added roughly 497,000 insured cow-years. 


About the Speaker
Honglin Li received his Ph.D. in Risk and Insurance from the University of Wisconsin–Madison. His primary research field is insurance economics. His research combines economic theory with large-scale administrative data to study how emerging technologies, insurance pricing regulations, and information frictions shape insurance market outcomes, including agricultural, health, and auto insurance markets.