2016年-世界发展银行全球_Knowledge_Externalities_from_Poverty_Mapping_in_the_European_Union_4页_726kb

报告摘要

Summary of Poverty in Europe (March 2016)

Core Content

The Poverty Mapping in the European Union (EU) project, a collaboration between the World Bank and the European Commission (EC), aimed to produce high-resolution poverty maps for EU member states. These maps use data from national population censuses and EU Statistics on Income and Living Conditions (EU-SILC) household surveys to estimate monetary poverty rates at the small geographic area level (e.g., counties, districts, municipalities). The project's goal was to improve the targeting of EU funds and inform national and sub-national policy decisions.

Main Views and Key Insights

Poverty in the EU: Over 122 million people in the EU are at risk of poverty or social exclusion, which is about 25% of the population. The Europe 2020 strategy aims to reduce this number by 20 million by 2020.
High-resolution Poverty Mapping: The project developed detailed poverty maps that provide more precise data than previous NUTS 2 level maps. These maps help identify regions with the highest poverty risk and enable more effective allocation of resources.
Methodological Innovations: The project led to the development of the Normal Mixtures-Empirical Bayes (NM-EB) method, which combines the strengths of both Empirical Bayes (EB) and Elbers-Lanjouw-Lanjouw (ELL) approaches.
PovMap Software: The World Bank's PovMap software was enhanced with these new methodologies, offering users more flexibility in choosing between ELL, EB, and NM-EB methods. This software is widely used for poverty mapping in low- and middle-income countries.
Data and Methodology Trade-offs: The choice between ELL and EB methods depends on two factors: the amount of data available and the degree of non-normality in the data. The EU has more comprehensive data, which suggests EB could be more suitable, but the size of the location effect is small, limiting gains.
Challenges in Poverty Estimation: Current methods struggle with accurately estimating poverty in the "tails" of the distribution, i.e., areas with very high or very low poverty rates. This highlights the need for further research in this area.
Knowledge Externalities: The project provided valuable insights into small area poverty estimation, influencing the World Bank's approach and promoting collaboration between researchers and national statistical institutes (NSIs).

Key Information

Data Sources: National population censuses and EU-SILC household surveys.
Geographic Disaggregation: Poverty maps are generated for small geographic areas, providing more detailed insights than NUTS 2 level data.
Software Development: PovMap 2.5 was updated to include more flexible estimation methods, including NM-EB.
Methodological Validation: A validation study compared ELL and EB methods across EU member states, leading to improved understanding and application of poverty mapping techniques.
Impact: The project not only met its original objectives but also had broader implications for poverty mapping in low- and middle-income countries.

Conclusion

The EU poverty mapping project significantly advanced the field of small area poverty estimation by introducing methodological innovations and improving the PovMap software. It provided a platform for technical exchange and collaboration between researchers and NSIs, and its findings are expected to benefit poverty mapping efforts globally, particularly in countries with limited data and analytical capacity.

Related Publications

Elbers, Chris and Roy van der Weide. 2014. "Estimation of Normal Mixtures in a Nested Error Model with an Application to Small Area Estimation of Poverty and Inequality." World Bank Policy Research Working Paper 6962.
Van der Weide, Roy. 2014. "GLS Estimation and Empirical Bayes Prediction for Linear Mixed Models with Heteroskedasticity and Sampling Weights: A Background Study for the POVMAP Project." World Bank Policy Research Working Paper 7028.

About the Team

World Bank Staff and Consultants

Kenneth Simler (Task Team Leader)
Robin Audy, Alexandru Cojocaru, Céline Ferré, Indiana Fonseca, Azhar Hussain, Maureen Itepu, Sandor Karacsony, Joost de Laat, Peter Lanjouw, Katarina Mathernova, Lei Pan, Ericka Rascon, Thomas Sohnesen, Roy van der Weide, Qinghua Zhao

National Statistical Institutes

Estonia: Julia Aru, Kaja Sõstra
Hungary: Judit Dobszayné Hennel, Éva Ménesi, Ildíkó Merkl
Latvia: Viktors Veretjanovs
Poland: Maciej Beresewicz, Tomasz Jozefowski, Tomasz Klimanek, Jacek Rowalewski, Anna Małasiewicz, Andrzej Młodak, Marcin Szymkowiak, Łukasz Wawrowski
Romania: Andreea Cambir, Nicoleta Caregea, Silvia Pisica
The Slovak Republic: Viera Doktoríková, Robert Vlăčuha
Slovenia: Danilo Dolenc, Tomaz Smrekar