OPHI Workshop on Robustness Methods for Multidimensional Welfare Analysis

  • Workshop programme
  • Participant biographies

The goal of this OPHI workshop is to identify and discuss specific areas for research aimed at developing stronger methodologies of robustness for multidimensional well-being analysis. Special emphasis is given to stochastic dominance techniques. The sessions will survey how various methods have been applied in the literature, and identify strengths and weaknesses of the techniques and discuss potential further developments. We see this workshop as a first stage, and hope that some participants and invitees might prepare papers for a special issue that follows up on key ideas.

The specific aims of the OPHI workshop are:

  1. To consider existing tests for the robustness and sensitivity of multidimensional comparisons when certain aspects of an index (e.g. weights, cut-off criteria) vary within reasonable bounds, and anticipate the development of additional tests.
  2. To consider the literature on stochastic dominance conditions and related tests, and discuss how and whether any other techniques should be extended to multidimensional space.
  3. As multidimensional welfare assessments become more and better established in the field, the need for robustness and sensitivity analysis increases accordingly. With the proliferation of composite indicators of general welfare and/or poverty or equality of opportunity, the robustness literature is concerned with the sensitivity of such indices to some of its key components like cut-offs (e.g. poverty lines, or criteria for defining multidimensional poverty), weights, equivalence scales, normalization criteria and other details that affect both identification (in the case of poverty indices, for instance) and aggregation stages. The OPHI workshop seeks to bring together both the available techniques and those currently under research in order to enhance multidimensional welfare analysis through the promotion of more and better robustness methodologies.
  4. There is a particular interest in stochastic dominance techniques among robustness methodologies. In recent years there have been breakthroughs in the development of both multidimensional stochastic dominance conditions and testing techniques that can only benefit the burgeoning multidimensional welfare literature. The workshop organizers feel that there is still scope for discussion, progress and dissemination of these techniques, which vary in aspects such as the completeness of the conditions they offer, the related testing toolkit and their practical implementation.