Governmental, public, and private organizations are more and more frequently required to make data available for external release in a selective and secure fashion. Most data are today released in the form of microdata, reporting information on individual respondents. The protection of microdata against
improper disclosure is therefore an issue that has become increasingly important and will continue to be so. This has created an increasing demand on organizations to devote resources for adequate protection of microdata.
In this chapter, we first characterize the microdata protection problem (in contrast to macrodata protection), discussing the disclosure risks at which microdata are exposed. We survey the main techniques that have been proposed to protect microdata from improper disclosure by distinguishing them in masking techniques (which protect data by masking or perturbing their values), and synthetic data generation techniques (which protect data by replacing them with plausible, but made up, values). We conclude the chapter with observations on measures for assessing disclosure risk and information
loss brought by the application of protection techniques.