GeoDevelopment Observatory of Bolivia / Tools
Where do lights predict poverty? GWR and MGWR
Geographically weighted regressions let the link between night-time lights and poverty vary across Bolivia.
This Colab notebook regresses multidimensional poverty on night-time lights per capita (2012) across the 339 municipalities. A single national regression explains 29% of the variation. A geographically weighted regression, with a bandwidth of 46 neighbours, raises that to 62%: every local slope is negative, but it is steepest in Cochabamba, Chuquisaca and Potosí and weakest in La Paz, Beni and Pando. A multiscale GWR lets the intercept and the slope vary at different scales and fits best of the three.