This webinar provides a high-level overview of a novel formal privacy framework, called per-record differential privacy (PRDP), that can be applied to business populations with highly skewed data. The research team, comprised of staff from the Census Bureau and Tumult Labs, has focused its initial work on the Census Bureau's County Business Patterns data product and have released the first demonstration table for public consumption.

This presentation will include the benefits and limitations of PRDP, its privacy guarantees, the determination and allocation of the privacy loss budget, as well as an overview of the demonstration table.