The Census data was at risk of exposing people’s private information. It was discovered that the current privacy settings, which involve switching users’ locations, are insufficient to guarantee user privacy. Additionally, this switching introduces bias in small areas, negatively affecting the accuracy of the analysis. The paper discusses the properties of the differential privacy method along with its advantages and disadvantages. It examines the balance between data accuracy and privacy protection. One advantage of differential privacy is its existing framework for indicating confidentiality measurements and its ability to disclose more data without risking the revelation of user identities.
Author:
the Population Reference Bureau and the U.S. Census Bureau’s 2020 Census Data Products and Dissemination Team
