All best practices on comparative surveys indicate that guidelines alone do not automatically ensure accuracy and comparability. Comparative statistical products depend crucially on process quality. The inventory of EU-SILC fieldwork practices presented in this paper shows that these processes vary enormously between Member States. Even nationally, optimal designs may thus fail to deliver comparability. The situation is aggravated by the fact that EU-SILC integrates several collections, one cross-sectional and several longitudinal of varying duration. They were designed to give answers to different questions, in particular measures of poverty at one point in time and sequences of poverty over time. If, however, the same cross-sectional indicators would be obtained from each component of EU-SILC they would be expected to give coherent results. Nonetheless, we observe discrepancies of hugely varying degree between Member States. In accordance with the ESS Vision 2020 this paper therefore argues for a new regime of “controlled flexibility” of harmonisation, including infrastructures which assist Member States in the design and control of their work.
This work has been supported by the second Network for the analysis of EU-SILC (NetSILC2), funded by Eurostat. The European Commission bears no responsibility for the analyses and conclusions, which are solely those of the authors.