A review of consumption poverty estimation for Mozambique

We broadly review the challenges encountered and choices made in the national assessments of consumption poverty using the 1996/97, 2002/03, and 2008/09 budget surveys. Efforts to maintain consistency with the previous survey imply that prior choices tend to be adopted in subsequent analyses. However, because 'best practice' evolves, there arises a natural tension between the desire to follow best practice and the desire to maintain consistency with previous analyses. New estimates produce qualitatively very similar results. We conclude that the principal conclusions of the poverty assessments are robust. The results also highlight the value of a consistent and reproducible approach.