Robustness of some EU-SILC based indicators at regional level

This Working Paper addresses some statistical aspects related to the construction from EU-SILC data of indicators of poverty and social exclusion for sub-national regions. Conceptual and methodological issues in going from the national level - for which EU-SILC surveys are primarily designed- to the regional level are discussed. Then five complementary approaches aimed at making the best use of available survey data for regional estimation are identified, and statistical considerations involved in their application are discussed. In addition to direct estimates from the survey, these include the construction of alternative measures utilising the available data more intensively, data cumulation over time, use of survey data in conjunction with external sources using small area estimation methods, and constructing supplementary regional indicators directly from administrative and other large-scale sources. The paper concludes with a plea for the provision of more complete information in EU-SILC microdata on sample structure and implementation and for the identification of regions in sufficient detail - information which would greatly enhance the usefulness of the data for regional analysis.