The impact of territorial factors on the total non-response error in the European Union - Survey on Income and Living Condition (EU-SILC)

This paper discusses a framework for the evaluation of accuracy of the Italian section of EU-SILC data with focus on non-sampling errors related to several components of the total non-response at household level. Following a classical hierarchical approach, classes of quality indicators are obtained by aggregating some ad hoc basic quality rates. Subsequently, in order to investigate the territorial perspective and its effects on the total nonresponse, one-way random effects ANOVA model and a multinomial logistic regression model are estimated. In this light, the work aims at exploring the main demographic and socio-economic factors potentially correlated with the survey participation at a subnational level. The final goal is sketching a territorial quality profile to identify those crucial areas with a more difficulty to be interviewed in order to define some ad hoc corrective interventions.