Immigration Article of the Day: Sanctuary and the Data State by Fatma Marouf & Huyen Pham

Sanctuary and the Data State by Fatma Marouf & Huyen Pham, Columbia Law Review, Vol. 126, Forthcoming 2026

Abstract
Historically, sanctuary policies focused on protected physical space and limits on police cooperation with Immigration and Customs Enforcement (ICE)—rejecting detainer requests, restricting jail access, and declining formal collaboration agreements. But as immigration enforcement has become increasingly mediated by integrated data systems, these visible forms of noncooperation have lost much of their practical force. Today, the scale and speed of immigration enforcement depend on the aggregation of data across federal agencies, state administrative systems, and commercial intermediaries.

Through expansive intergovernmental data sharing and purchases from private data brokers, ICE has embedded itself in information streams far removed from traditional law enforcement, including motor vehicle records, public benefits databases, biometric systems, and consumer geolocation data. Ordinary administrative and commercial information now function as enforcement infrastructure. 

This Essay argues that sanctuary must therefore be reconceptualized as a problem of information governance. A critical dimension of sanctuary lies in how states and localities design rules governing the collection, retention, sharing, and aggregation of data about their residents. Drawing on an original dataset of state and local laws, the Essay develops a framework for “data sanctuaries” that operate through data minimization, purpose limitation, access controls, broker regulation, and aggregation constraints. Rather than displacing traditional place-based or policing-focused protections, this framework adds an informational layer that regulates the data systems through which modern immigration enforcement operates.