Daniel Gerszon Mahler, Jose Montes, and David Newhouse Target 9.c of the Sustainable Development Goals calls for the achievement of universal and affordable internet access by 2020. This note analyzes Sub-Saharan Africa’s progress towards this goal. It finds that (i) rates of internet access reported in household surveys differ markedly and are often lower than the prevailing estimates of internet use reported by the International Telecommunications Union, (ii) internet access in regions outside the capital city tends to be lagging and, (iii) lack of access to electricity is a key barrier constraining access to internet among poor Africans. Figure 1: Internet usage in Sub-Saharan Africa Access to internet is essential for businesses, public institutions, and households to flourish in the (a) Internet usage in SSA and the rest of the world modern economy. In the private and public sector, internet access can help spur productivity gains and deliver services more efficiently. For households, internet access can increase opportunities, build human capital, connect households to other parts of the country, and contribute to personal well-being. Yet Sub-Saharan Africa remains a long way from achieving universal internet access. According to the International Telecommunications Union (ITU), which tracks internet usage globally and across countries, only 1 in 5 in Sub-Saharan Africa used the internet in 2017. While internet access in Sub-Saharan Africa has (b) Internet usage by country in SSA, 2017 grown rapidly in recent years, access rates remain well behind the rest of world (Figure 1a). Internet usage differs markedly by country within Sub-Saharan Africa (Figure 1b). Whereas more than half the population uses the internet in South Africa, rates are closer to 30% in West Africa, and only around 10% in Central Africa. Internet usage is particularly low in landlocked countries, where the physical infrastructure necessary to provide infrastructure is costlier, and access is also more dependent on neighboring countries. Source: International Telpecommunication Union (ITU), World Telecommunication/ICT Development Report and database. To analyze the people and places that are lagging in Table 1: Household surveys with comparable data on internet access the digital revolution in greater detail, microdata Share of Share of from household surveys are needed. SSAPOV, a population with population Survey database of harmonized nationally representative Country year internet access using the in their home internet household surveys in Sub-Saharan Africa, contains (SSAPOV) (ITU) harmonized data on internet access and many other Benin 2015 2% 11% variables. Although not all household surveys have Burkina Faso 2014 1% 9% questions on internet access, the ones that do can be Burundi 2013 0% 1% utilized to better understand the types of households Cameroon 2014 5% 16% that lack access to internet. Since 2010, nearly half of Chad 2011 10% 2% the countries in Sub-Saharan Africa have conducted Comoros 2013 2% 7% a household survey with comparable information on Congo, DR 2012 2% 2% internet access, as shown in Table 1. Ghana 2012 8% 11% Kenya 2015 27% 17% The measure of internet access contained in SSAPOV Madagascar 2012 1% 2% is different from the measure tracked by the ITU. Mauritania 2014 3% 11% Mauritius 2017 56% 56% Whereas the former is concerned with internet Namibia 2015 15% 26% access, the latter is concerned with internet usage. Niger 2014 6% 1% Internet users – as defined by the ITU – are individuals Rwanda 2016 17% 20% who have used the internet from any location in the Senegal 2011 4% 10% last 3 months. This includes using an internet- Seychelles 2013 37% 50% enabled computer, mobile phone, video game Sierra Leonne 2011 1% 1% console, digital TV, or other internet-connected South Africa 2010 7% 24% device. In contrast, internet access as defined in Tanzania 2011 1% 3% SSAPOV implies that households have an internet Uganda 2016 14% 22% connection within their homes. Although the two are Source: SSAPOV database, Sub-Saharan Africa Team for positively correlated, as evident from Table 1, the Statistical Development, World Bank, Washington DC and differences between the two measures can be large. International Telecommunication Union (ITU), World Telecommunication/ICT Development Report and database. Furthermore, in some countries like Chad, access in SSAPOV substantially exceeds the usage rate Because internet use is growing rapidly, we analyze according to the ITU. Aside from the different six recent surveys carried out since 2015 with concepts the two measures are trying to capture, information on internet access: Benin (2015), Kenya discrepancies such as these are also caused by (2015), Mauritius (2017), Namibia (2015), Rwanda differences in data sources. SSAPOV relies on (2016), and Uganda (2016). These countries both nationally representative household surveys, while span Sub-Saharan Africa and represent low-income ITU’s methods are less transparent; the ITU either countries, lower-middle-income countries, and estimates usage rates themselves or obtains upper-middle-income countries. information from questionnaires filled out by NSOs or other national agencies, who in turn may obtain Because the surveys in SSAPOV are the same ones data from a variety of sources. that are used to measure poverty, they are well- suited to explore the digital divide between poorer and wealthier households. Unsurprisingly, in all six the steep gradient. In Kenya, for example, less than countries internet access is substantially higher for 5% of the poorest decile had access to internet in better-off households with higher per capita 2015, while 2 in 3 of the wealthiest decile did. consumption (Figure 2). What is more surprising is Figure 2: Internet and electricity access by consumption level Source: Benin Enquête Modulaire Intégrée sur les Conditions de Vie des Ménages (2015), Kenyan Integrated Household Budget Survey 2015- 16, Mauritius Household Budget Survey (2017), Namibia National Household Income and Expenditure Survey, Rwanda Integrated Household Living Conditions Survey 5, Uganda National Household Survey (2016/17). Note: Consumption levels below the 1st percentile and above the 99th percentile are not plotted. Electricity access in Mauritius is assumed to be universal. According to the Sustainable Energy for all initiative, about 99% of Mauritians have electricity access. In most countries, electricity is a key constraint to necessary infrastructure is less profitable for internet internet access for poor households. The exception is providers. When looking closer at the spatial Mauritius, which has near universal electricity access. distribution of internet access, in many countries, For the bottom 40 percent of the other five countries, only the capital region has high levels of internet only between 3% and 21% of those that lack internet access while other regions tend to lag. The low rates access have electricity access. The households that of reported access outside the capital highlight the lack both internet and electricity face at least two importance of expanding the availability of internet large impediments to be connected, proper to secondary cities and towns. infrastructure and the resources to purchase a device with access to the internet. To ensure that gains in internet access reach the poor going forward, it is fundamental to better understand A substantial portion of better-off households in all what governments in Sub-Saharan Africa are doing six countries report access to electricity but no to expand access to both electricity and internet, internet. The share of the top 60% in this category especially outside of capital cities. The World Bank ranges from 21% in Uganda to nearly 50% in can make an important contribution by documenting Namibia. For these households, the barriers to these efforts and systematically utilizing nationally internet adoption could include computer illiteracy representative household surveys to track their and high costs of internet services, which potentially success in expanding access to the poor. stem from ineffective competition policies, regulation, or the geographical location of households. ABOUT THE AUTHORS Daniel Gerszon Mahler is a Young Professional at The latter can be analyzed by disaggregating internet the World Bank’s Poverty and Equity Global Practice access by location. Rural households, which on (GPV). dmahler@worldbank.org average are poorer in all six countries, also face lower rates of internet access. The urban-rural gap in access Jose Montes is a Data Scientist at the World Bank’s is widest for better-off households. In Kenya and Poverty and Equity Global Practice (GPV). Uganda, rural households at the 90th percentile of jmontes@worldbank.org the national distribution have the same probability of David Newhouse is a Senior Economist at the World having internet access as urban households living at Bank’s Poverty and Equity Global Practice (GPV). the international poverty line. Rural households may dnewhouse@worldbank.org face lower rates of internet access because their geographical location implies that building the This note series is intended to summarize good practices and key policy findings on Poverty-related topics. The views expressed in the notes are those of the authors and do not necessarily reflect those of the World Bank, its board or its member countries. Copies of these notes series are available on www.worldbank.org/poverty