This research examines the utility of differential privacy in preserving the privacy of published data from randomized control trials (RCTs). The study uses simulation studies and replication packages in economics and policy analysis to assess the impact of differential privacy on the analyses. The findings indicate that differential privacy-based methods can provide valid protection for the published data. However, computational issues may limit the use of these methods for more complex analyses. The results are particularly relevant for researchers who wish to share randomized control trial (RCTs) data with strong privacy protection.
Publication date: 28 Sep 2023
Project Page: ?
Paper: https://arxiv.org/pdf/2309.14581
