Examining Digital Connectivity in Kenya's 2019 Census Data
For more detail on this project, visit the resources linked below:
Digital technologies are rapidly changing the way economies operate. Governments across the world are increasingly prioritizing increasing their citizens' access to digital tools as part of their economic development plans. This is particularly true in developing countries — for instance, in Kenya (where I am from) the government intends to make the country “a nation where every citizen, enterprise, and organization has digital access and the capability to participate in the digital economy”.
However, for governments to efficiently invest in digital infrastructure, they need to understand their citizens’ current levels of connectivity. Kenya's most recent 2019 census surveyed the population’s mobile phone ownership and internet usage. In this project, we examine this data to understand the level of digital connectivity in Kenya.
Why digital connectivity matters?
Understanding digital connectivity levels can help the Kenyan government identify regions that require greater investment in digital infrastructure. Moreover, this information can also inform strategy for other digital service providers (e.g., financial services, agriculture advisory).
Research Questions: Using R and the Shiny library I investigated the following questions:
- What percentage of Kenya’s total population is digitally connected? *
- How does digital connectivity vary across different regions of the country?
- How does digital connectivity vary between genders?
- What is the level of digital connectivity in areas where different forms of agriculture are practiced?
Note: * For this project, “digitally connected” was defined as owning a mobile phone and/or internet usage.
Digital Connectivity Explorer
We visualized mobile phone ownership and internet use across Kenyan counties using chloropleth maps built using the leaflet package in R. (To view the interactive charts, visit the Shiny App linked at the top of this post). Below are some of the takeaways:
- Mobile phone ownership was highest (50-70%) in the central and southern counties of the country - particularly around the capital city Nairobi - with pockets of moderate phone ownership (40-50%) in the western counties.
- Internet usage was lowest (below 10%) in north and northeastern counties. The World Bank reports that these regions are characterized by profound infrastructure deficits, including lack of access to roads, electricity, water, and to social services. The average poverty rate in these regions is double the national average (68% vs 34%).
The relationship between phone ownership, internet use, and population size
- As expected, higher mobile phone ownership correlated with higher internet use. With a few exceptions, larger population centers (shown by larger circles in chart below) had the highest connectivity (see Nairobi in the top-right). The counties in the bottom left quadrant of the chart (i.e., low phone ownership and low internet usage) were predominantly counties in the north and northeastern parts of Kenya.
- The lack of digital connectivity in Kenya's northern counties likely has broader implications on these communities' ability to participate in the digital economy.
Mind the (Gender) Gap
At the national level, men tended to have marginally higher access to mobile phones and internet than women. However, when we examine the connectivity data at a county level, we see a larger digital divide between the genders.
In the bar chart above, we observe that a higher proportion of men owned phones in 29 out of 47 (61%) of counties. In some places, men's phone ownership rate was 8% higher than that of women. The inequality in internet usage was even starker - men reported using the internet more than women in every single county.
Mobile phone ownership in areas where agriculture is practiced
Agriculture is one of the most important drivers of Kenya's economy, contributing approximately 33% to the country's GDP and employing approx. 60% of the population (World Bank). Digital agricultural services ( e.g. financial services, agricultural advisory, market access) are increasingly being used to boost the sector's productivity, with Kenya at the forefront in Africa.
The maps below compare mobile phone ownership rates against the number of households participating in different types of agriculture, specifically subsistence farming, commercial farming, and cattle rearing.
Mobile Phone Ownership in areas with subsistence farming households
Central counties had the highest overlap between high proportion of subsistence farming households and high rates of phone ownership. In contrast, parts of western Kenya and the northeast have a relatively high proportion of subsistence farming households but lower phone ownership.
Mobile Phone Ownership in areas with commercial farming households
The central regions of the country, where the proportion (%) of commercial households is higher, tend to have higher rates of phone ownership. However, parts of western and eastern Kenya that have a relatively high proportion of commercial farming households have much lower phone ownership.
Mobile Phone Ownership in areas with cattle-rearing households
Cattle rearing is prevalent in the northern areas of the country, where phone ownership is low. This highlights the lack of digital infrastructure in these regions relative to other parts of the country.
What have we learned?
- Digital connectivity is not equally distributed in Kenya. It is highest in central counties and pockets of western Kenya; it is moderate in the east and lowest in the north and northeast
- National statistics obscure the gender gap in digital connectivity. The digital divide is most apparent at county level, with men having greater access and usage.
- Lower digital connectivity in the northern and western counties might hinder agricultural household's economic productivity.
- Examine agricultural output data across counties to investigate the possible economic implications of low digital connectivity.
- Explore agricultural household data by gender at the county level to get a better understanding of the relationship between digital connectivity and agricultural productivity. The census data did not include this breakdown