Synthetic Data and Public Policy: supporting real-world policymakers with algorithmically generated data

Abstract

Good policy is best developed by drawing on a wide array of high-quality evidence. The rapid growth of data science and the emergence of big datasets has materially advanced the supply and use of quantitative evidence. However, some key constraints remain, including that available datasets are still not big enough for some analytical purposes. There are also privacy and data security risks. Synthetic data is an emerging area of data science that can potentially support policy decision making through enabling research to work faster and with fewer errors while also ensuring privacy and security.