Alternative Data
Data not traditionally used in underwriting, such as credit history, driving behavior, or digital signals. Often regulated as ECDIS when applied to consumers.
Alternative data, also called non-traditional data, is information used in insurance decisions that does not come from standard application questions or medical exams. Examples include credit history, telematics driving data, wearable-device data, online behavior, and purchase patterns.
Alternative data can improve risk assessment, but it also raises fairness and privacy questions. A variable that is predictive of risk may also be a proxy for race, income, or geography. When alternative data is obtained from third parties about a consumer, it often falls under the regulatory category of external consumer data and information sources (ECDIS).
Carriers should document why each alternative data variable was selected, test it for disparate impact, and ensure they have the legal right to use it. See our glossary entry on external consumer data and information sources and guides to Colorado SB 21-169 and NYDFS Circular Letter No. 7.