Introduction:
Federal statistical agencies have longstanding concern over confidentiality of their data (sample surveys and censuses). Both the identities of data subjects and sensitive attributes in the data must be protected [1, 2]. But the agencies also have an obligation to report information to the public.
This tension between confidentiality and dissemination of statistical information [3] arises equally sharply in non-governmental contexts, from electronic medical records [4] to E-commerce transaction data.
Confidentiality is threatened by advances in information technology, such as powerful capabilities for record linkage across multiple databases. Other new technologies, however, not only protect confidentiality, but also meet user needs in innovative ways. Here we describe a system (see www.niss.org/dg) being developed by the National Institute of Statistical Sciences for the National Agricultural Statistics Service (NASS), which disseminates survey data on usage of agricultural chemicals in far greater geographical detail than previously, but protects the identities of farms in the survey
tr107.pdf