Local Identities of Poverty: poverty narratives in decentralized government and the role of poverty research in Uganda.

This paper takes as its point of departure recent research that suggests propoor policies are more likely to be supported by central elites than local elites, and that, in contexts where there is significant decentralization of government, central government may need to challenge local elite resistance to ensure effective implementation of those policies. The paper explores whether poverty reduction strategies can be made more effective use of research to effect a ’reidentification’ of the nature and causes of poverty. It focuses on Uganda, where research has had a high profile in the government’s Poverty Eradication Action Plan, particularly through the Uganda National Household Survey (UNHS), and the Uganda Participatory Poverty Assessment Programme (UPPAP). The paper examines the methodology and impact of these research programmes. It argues that the UNHS could, with appropriate changes in how expenditure is interpreted in terms of well-being, be useful as an indicator of national poverty trends, but sample structure constrains its application as a means of monitoring decentralized delivery of povertyreduction programmes. The current UPPAP methodology provides a rich source of detail on the nature of poverty in Uganda, but the validity of its findings is undermined by emphasis on aggregation of local information to national level with consequent loss of contextualised understanding of social processes creating poverty. The paper argues that, in illuminating such processes, a more rigorous application of qualitative research method would challenge assumptions about the 'residual' nature of poverty that underlie both 'participatory research' and decentralization as a strategies of poverty reduction