Page 331 - Contributed Paper Session (CPS) - Volume 6
P. 331
CPS1949 Khalid S.
3. Summary of the results: measurement indices of poverty according to
approaches of measure and weighting schemes
By adopting AF method, the results show that all indices of poverty
(headcount ratio of poverty, multidimensional poverty index (MPI),
vulnerability and severe poverty) estimated according the AF weighting
scheme are significantly higher than those estimated using the other pre-
determined normative weights. The stochastic dominance of the curves of
poverty allows us to confirm the robustness of these results.
Graph 1: Multidimensional indices of poverty according to AF method : sensibility to
weighting schemes. Factors structured by dimension (%)
Source: Author’s calculation
If we don’t structure the items by dimension we’ll have an overvaluation of
the index of multidimensional poverty, in particular when using the AF
weighting scheme. This tendency concerns all the weighting schemes, the
headcount of poverty and the multidimensional poverty index (MPI). In light
of these results, it turns out important to rethink the broad configuration of
the deprivation items by dimension.
Graph 2: Multidimensional indices of poverty according to the approach AF Factors
aren’t structured by dimension
Source: Author’s calculation
By structuring the items of deprivation by dimension, the results of the
fuzzy logic approach show that, regardless of weighting scheme, the indices
320 | I S I W S C 2 0 1 9