Ever wanted to see a complete and honest view of the “aid world” for an entire country?
Using data from the GlobalGiving Storytelling Project in East Africa (about 65,000 stories from people, collected in 2010 and 2011), my buddy Nick (co-creator of aidsight.org) was able to build me this genetic tree of 2,558 named organizations in Kenya and Uganda. The algorithm takes stories told by citizens in each country about one organization and groups organizations together that have similar stories. The map reveals the many aid sectors that do the work.
This is useful because allows me to evaluate the success and failure rates of organizations by the type of work they do. I’ll present results on that in a future post. First, I’ll show some interesting work clusters it found, and then post a link the full map, which is best viewed offline.
The condensed, annotated version
(presented in no particular order)
Local self-help groups (there are tons of these)
Faith-based local organizations
Child welfare, rescue and protection organizations
Youth groups, which are closely related to women groups, which include some co-ops and self-help groups. The focus here is on the member identity, and not on the “self-help” part
Health organizations (a massive tree)
Income generation (along with some other odd stuff)
Rights-based organizations focusing on children and labor
Education, trade associations, and development
What this view of aid doesn’t show are the overwhelming power that international organizations have over the rest. They are just one name among thousands.
Here is the full map, for download: download the full image
This map was created using scikit learn’s cluster.hierarchy.dendrogram algorithm.