Small area estimation methods for poverty mapping: a selective review Articles uri icon

publication date

  • January 2019

start page

  • 11

end page

  • 22

issue

  • 1

volume

  • 17

Electronic International Standard Serial Number (EISSN)

  • 2452-7395

abstract

  • Poverty mapping in small areas is currently having increasing interest, because those mapsaid governments and international organizations to design, apply and monitor more effectivelyregional development polices, directing them to the actual places or population subgroups wherethey are more urgently needed. After a simulated census method used by the World Bank, severalother procedures have been developed that proved to have better properties. We will review severalmethods that are applied for poverty mapping in small areas, including those based on area levelmodes and used by the U. S. Census Bureau for estimating poor school age children and methodsbased on unit level models such as the traditional method used by the World Bank and empiricalbest (EB) and hierarchical Bayes (HB) methods based on optimality criteria. We will also discusssome variations of the unit level model methods that can used to deal with certain situations suchas informative sampling or two-stage sampling. We will discuss pros and cons of these methodsfrom a practical point of view, but based on the theory that is currently known.

keywords

  • area level model; empirical best estimation; hierarchical bayes estimation; local; poverty indicators; unit level models