Analysis of Multidimensional Poverty (PDF)

This book presents a new method for measuring multidimensional poverty. The author critically analyzes various statistical approaches, and proposes a new way of applying a factorial technique, Multiple Correspondence Analysis, to poverty analysis. The core of this new approach rests on the identification of poverty types and on the construction of a Composite Indicator of Poverty, which objectively weights the multiple categories of poverty. Built as a two-step approach, Multiple Correspondence Analysis allows for the integration of unidimensional money-metric techniques. The author reflects on the conceptual debates and dimensions of poverty, presenting empirical studies showing the strengths and weaknesses of various poverty-measurement methodologies. The first part of the book is theoretical. In it, the author uses examples to develop the rationale underlying the new approach. The second part of the book presents two case studies: one from Vietnam and the other from Senegal. Each study uses Multiple Correspondence Analysis and both are based on large household surveys implemented by the respective national statistical offices. The unique structure of this book makes it practical for use by poverty and policy researchers, professionals in international development, and graduate students interested in poverty and inequality.