Factoring Local Priorities into Sustainable Development Index

[jgdenby]

Thanks for presenting!

In the PNAS paper, you define a sustainable development index X_i in terms of the fraction of population in a given spatial unit i with access to water, electricity, sanitation, and homes – four components of Secure Tenure Index (STI) – with more parameters to come as data concerning other components are collected.

I wonder, as this measure increases in complexity with the inclusion of additional STI components, whether there is some use for relative weighting of those components, either a priori or based on results from surveys of local priorities, like those reported in Table 1? For example, proportion p of population i having access to clean water might exert more influence on the index measure than proportion p of i having access to education, either because UN-Habitat has standardized a relative importance hierarchy or because local surveys indicate such a priority. If weighting were included, the index might more faithfully characterize a region by measuring how well its population is able to meet its specific needs.

[tamos]

I really like the idea of refining the weighting based on surveys. This idea is very present in the development/humanitarian sector.

[lmbett]

Good question (we also got it from reviewers). You could/should consider weighting, but that make the index more complex so you'd better have a good reason ... The index can always be decomposed (in a self-similar way) into components, so we thought it was not very necessary... but could be done.