Workshop on Value Judgements in Multidimensional Poverty Measurement Design

Research Workshop | 28-29 June, Oxford Department of International Development, Oxford 2012

This intensive workshop joined together philosophers and economists to discuss a very specific, practical topic – the design of new multidimensional poverty measures – and flesh out some of the practical issues that arise in so doing, namely how to use insights from statistical and normative reasoning in the choice and justification of indicators, weights and cutoffs in a poverty measure, and what kind and quality of public debate or deliberative discussion is ‘enough’ for national official poverty measures.

Brief description of topic

Innovative national and international quantitative multidimensional poverty measures are being used as official poverty measures and for policy design and monitoring. However many decisions in the design of these multidimensional poverty measures entail value judgements. These include value judgements about what poverty is (dimensions, indicators and weights), and who is poor (cutoffs). Because resulting measures must also fulfil technical standards, their design requires an epistemological framework to coordinate the use of empirical and normative considerations. The design and defence of normative judgements is relatively underdeveloped and dispersed. Furthermore, economists and statisticians are not trained in value-judgements so often do not address these issues. Yet it is the ethical choices in multidimensional poverty measurement that become the most controversial in public policy and in the media. The proposed research workshop will engage philosophers and social scientists on certain specific, pointed and practical questions that arise in measurement design. The following questions are answered in any multidimensional poverty measure: our workshop will consider how such questions can best be answered in the design of official national multidimensional poverty measures, articulating diverse considerations and generating practical and non-technical as well as academic outputs with respect to:

  1. Purpose & Space – What is the poverty measure for? In which space ought it be constructed? [most use resources due to data constraints; how justify when these proxy functionings]
  2. Dimensions – How should the ‘dimensions’ of poverty be selected (e.g. health, work)?
  3. Particular Indicators – How should particular indicators of poverty be chosen? What is the role of statistical vs normative insights?
  4. Cutoffs – How to decide ‘how much is enough’ in each indicator?
  5. Weights – How to set and justify relative values or weights on indicators?
  6. Procedure – Who decides normative issues? What is the appropriate role of poor people, governments and statistical or technical experts?
  7. Plural Criteria – How should statistical, political, and participatory input be coordinated in measurement design?