2017 Articles

The natural and capital infrastructure of potential post-electrification wealth creation in Kenya

Ponce de Leon Barido, Diego; Fobi Nsutezo, Sally Simone; Taneja, Jay

Background:
It is widely accepted that electricity is an important element for improving levels of human development and wealth creation in rural areas. Yet, little research has explored the conditions under which electrification could lead to wealth creation post-electrification. Using Kenya as a case study, this paper uses natural capital (NC) and infrastructural capital (IC) data to compare the enabling environments under which electrification could lead to wealth creation (and persistent demand for electricity) post-electrification.


Methods:
We use multiple spatial data sets to create three different metrics for NC and IC and use them to create a micro-enterprise development index (MED index). NC data is composed of water body data (major rivers and access to irrigation infrastructure), soil data (soil quality and agro-ecological potential), and agricultural data (crop intensity and diversity). IC is composed of spatial data spanning major towns, first and second tier roads, electricity infrastructure (transmission grid, location of government, and entrepreneur run off-grid electrification projects), population density, access to education, trade centers (markets), healthcare, and access to financial services. We perform feature scaling on NC and IC data and use them to create a MED index, which we use to represent the potential for rural micro-enterprises to create wealth, post-electrification. We compare this spatial proxy to a nightlights GDP per capita proxy developed by the World Bank in 2015 and provide a discussion highlighting the benefits and drawbacks of our approach and of using nightlights as a single metric for wealth in rural areas.


Results:
In Kenya, infrastructural capital follows natural capital. Regions with greater natural capital have relatively higher development and penetration of infrastructural capital. We observe a large discrepancy between our MED index and the nightlights income proxy, which could be caused by an underestimate of economic activity by nightlights, and an overestimate by the MED index, being that it is a measure of “potential” wealth (as opposed to current wealth).


Conclusions:
A spatial aggregation of natural and infrastructural capital, and nightlights data, could be an accurate demand-side input for electrification supply-side models including grid-expansion and off-grid strategies.