Strengthening the Production of Public Use Microdata Files: Better Tools for Anonymizing Census and Survey Data

Public use microdata files (PUMFs, PUFs or many other acronyms) are the bread and butter of many data centres and are essential tools for research and teaching. To date, the production of such files has been a challenging and time-consuming task for data producers. This makes it an expensive step which is often not undertaken by data producers. Intuitively, the anonymization of data files should lend itself to the use of computer based tools to facilitate this process. The International Household Survey Network (IHSN) has established a task force which is investigating the development of a series of tools to aid data producers in measuring the disclosure risk associated with a file, identifying ways of reducing this risk and assessing the information loss as a result of implementing the disclosure limitation procedures. Rather than an integrated software tool, the task force is working towards a modular, coherent set of tools to achieve the above goals. By building on work by others, they may even be able to develop an "intelligent" system. This presentation provides an overview the work of this task force.