Estimation of poverty indicators for domains

Risto Lehtonen, Ari Veijanen

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review


We present results obtained in the context of the research project AMELI (Advanced Methodology for European Laeken Indicators) on the estimation of poverty indicators (Laeken indicators) for population subgroups or domains and small areas. The indicators discussed are at-risk-of poverty rate, the Gini coefficient, poverty gap and quintile share ratio. We use design-based direct estimators, generalized regression estimators, model calibration estimators, empirical best predictor type estimators, model-based prediction-type estimators and composite estimators. New prediction-type estimators are introduced. The methods use often unit-level auxiliary data. We develop a frequency-calibrated estimator for cases where only aggregate auxiliary data are available. We also discuss the case of outlier contamination. Results are based on simulation experiments where we use register data maintained by Statistics Finland.
Original languageEnglish
Title of host publicationNTTS 2011; Paper 2 session 14
Number of pages10
Place of PublicationBrussels
PublisherEuropean Commission
Publication date2011
Publication statusPublished - 2011
MoE publication typeA4 Article in conference proceedings
EventNTTS 2011 Conference on New Techniques and Technologies for Statistics - Brussels, Belgium
Duration: 21 Feb 201125 Feb 2011

Fields of Science

  • 112 Statistics and probability

Cite this