Poverty, Poverty Measurement and the Capability Approach

At the German Development Institute / Deutsches Institut für Entwicklungspolitik (DIE) the multidimensional “Correlation Sensitive Poverty Index” (CSPI) has been developed which, in contrast to the well-known “Multidimensional Poverty Index” (MPI), accounts for the fact that the poor differ with regard to the severity of deprivations they face. Against this backdrop, this project examined, amongst other issues, how the way we measure poverty affects what we actually think to know about it, what the possible policy implications might be and how measurements can be improved.


Project Team:
Nicole Rippin

Time frame:
2010 - 2013 / completed

Co-operation Partner:

Department of Development Economics, University of Göttingen

Project description

Amartya Sen's Nobel Prize-winning Capability Approach revolutionized poverty measurement by challenging the traditional way of measuring poverty in terms of insufficient income. In fact, the income approach has increasingly been criticised as especially the poor have quite often little opportunity to monetarily satisfy their needs, in particular in case of public goods like health or edution. The capability approach overcomes this problem by measuring different kinds of deprivations directly. It found its way into international politics with UNDP's introduction of the "Multidimensional Poverty Index (MPI)" in 2010. The MPI, however, does not account for the fact that the poor differ with regard to the severity of deprivations that they face. In alignment with Amartya Sen who stresses the particular importance of distribution-sensitivity in poverty measurement, the multidimensional and distribution-sensitive "Correlation Sensitive Poverty Index" (CSPI) has been developed at DIE and inspired several questions:

1. How should poverty be measured in the future? How do extent and shape of global poverty change according to different poverty measures? What are the policy implications?

2. Considering the very limited availability of data, which dimensions and indicators can and should composite indices like MPI and CSPI utilise? And which weights? What are the main missing dimensions?

3. Which questions would have to be included in future household surveys in order to be able to capture the main missing dimensions? How should questions in general be formulated in order to be able to provide not only information about achieved functionings, like being well-nourished (which includes starving and fasting), but capabilities, i.e. the ability to be well-nourished (though one might make a different choice)?

4. The capability approach is considered to provide a convincing argument for the existence of poverty even in affluent countries. What does poverty look like in affluent countries like Germany? What are the main differences with regard to developing countries? And what can we learn from the experience of much more advanced data sets like the German Socio-Economic Panel with regard to the inclusion of missing dimensions, choice of indicators and the formulation of capability questions?

Publications