What Studies in Spatial Development Show in Ethiopia-Part II

Priyanka Kanth's picture

In Part I of our blog —based on a background note we wrote for the World Bank’s 2017–2022 Country Partnership Framework for Ethiopia—we presented our key findings on the spatial or regional distribution of poverty and child malnutrition in Ethiopia.

In Part II of our blog, we look at changes in road density over the ten years from 2006 to 2016, and in nightlights in six cities over four years from 2012 to 2016.

Changes in road density pointed to greater economic concentration towards the center of Ethiopia and the north of the country. These are also areas of greater population density. Figure 2a shows that, between 2006 and 2016, the increase in road density was concentrated in certain regions, notably Ethiopia’s capital of Addis Ababa, as well as Tigray in the north of the country and in Oromia in the center.

Figure 2a: Changes in road density and length between 2006 and 2016

Source: World Bank visualization based on data from various UN agencies

Figure 2b: Rural Access Index (RAI) and major roads in 2016

Source: World Bank visualization based on data RAI (World Bank).

Remote and economically lagging regions, and Amhara Region, see lesser increases in road density. Taking the development of roads as a proxy for the development of infrastructure, this suggests that infrastructure development has not been homogeneous across all regions. It also shows that road connectivity for some regions is poor, both within those regions and with other regions, with consequences for labor mobility, the transportation of goods and services, and for agricultural productivity as the distance and travel times to markets are longer.

Despite the large infrastructure investments undertaken by the Ethiopian government in the past ten years, accessibility by road to rural areas remains low in Ethiopia; we can see its distribution across the country in Figure 2b. The Rural Access Index was 21.6 percent in 2016, signifying that only around 22 percent of the rural population had access within a 2km distance of them to a “decent” road.

Twinkle, twinkle little light

Finally, we look at nightlights in some of the secondary cities. From Figure 3 it could be interpreted that urban GDP is stagnant as there were no significant changes in the density or distribution of night-time lights over the period of 2012–2016, even though urbanization outside Addis Ababa was ongoing and urban poverty has been reduced since 2012.

As per the World Development Report 2009, nightlights are not a good proxy for GDP; however, differences in the pattern of nightlights over a given time are correlated with changes in GDP. In figures 3a-b, we see that the trend of nightlights across several secondary cities in Ethiopia remains constant. Therefore, secondary cities don’t seem to grow in keeping with Addis Ababa, the largest urban center of the country.

Figure 3: Spatial dimension of nightlights
a) Total brightness of nighttime lights

Source: World Bank visualization based on data from NOAA’s VIIRS Satellite.

b) Average brightness of nighttime lights.

Source: World Bank visualization based on data from NOAA’s VIIRS Satellite
The significant increase in GDP witnessed by Ethiopia is primarily due to growth in agriculture in rural areas and in the service sector. The country’s push for developing its manufacturing sector is relatively recent and might explain the figures for secondary cities better. It is likely that secondary cities are witnessing growth in the service sector—and not in industrial or manufacturing sectors—and that this growth is therefore not resulting in any significant changes in the number or density of nightlights.

In addition, the influx of migrants into Addis Ababa is higher than in other secondary cities, suggesting that real and perceived opportunities lag behind in secondary cities.

Combining our analytical findings with our visualization of spatial development, we concluded that spatial development outcomes could be increased through interventions, primarily in four areas, which will we explore in Part III of our blog series.


Data provided by DEC survey unit. Maps produced by the Data Management Unit.