In this seminar, we discuss with interested parties the conditions necessary to implement the SynLBD approach, with the goal of providing other statistical agencies a straightforward toolkit to implement the same procedure on their own data. Our hope is that by implementing similar procedures on comparable business microdata, new research both within and across countries can be enabled. The ideal end result is a series of country-specific datasets on establishments and/or firms available within the same computing environment. We discuss the data and software requirements for the lowest-cost approach, the disclosure protection statistics already implemented that can be used to achieve release of the data in this way, the validation procedures that an agency should agree to, and the likely cost of maintaining such procedures. The seminar brings together academics working on cutting-edge methods for the protection of privacy in statistical databases, and researchers and implementers at statistical agencies that have started or are interested in starting a similar project.
Five sessions will touch on the full lifecycle of a SynLBD development and implementation, and will follow the same pattern. We will first discuss existing implementations and experiences, and will then as a group discuss issues as they pertain to the broader community. Emphasis should be on discussing open issues, specific solutions to specific problems.
For more information, a link to the program and to register, please consult our calendar page.
Funding for the workshop is provided by theNational Science Foundation (CNS-1012593, SES-1131848) and the Alfred P. Sloan Foundation. The organizers thank the National Academies’ Committee on National Statistics for hosting the seminar.