On 5 October 2016, the Ethiopian railway corporation launched a 750 KM rail-line connecting the landlocked country from its capital, Addis Ababa, to Djibouti, its strategic economic link to global commerce. A few hours later, the communication ministry completely shut down all Internet connectivity across the country, with the stated aim of quelling protests in parts of the country. Spending millions of dollars to connect a country to the world through a railway, while intentionally shutting down the country’s Internet connectivity on the same day is a quite a paradox. To consider a whole city, or even a country, intentionally disconnected off the Internet for days by their government, may sound quite abstract, but more than fifty incidences like these were recorded globally in 2017, of which for every two of these, one was happening in Africa.
The effects of these intentional Internet disruptions have ranged from increased citizenry backlash, economic losses, and eroded international reputation. What is interesting though, as seen from the Ethiopian vignette above, is how disrupting the Internet contradicts the very economic plans of such countries. On the one side, countries are investing heavily on communication and transport infrastructure for economic connectivity yet easily reversing the marginal gains made by their intentional Internet disconnections.
Today we are releasing findings from our continuing research on Internet disruptions, together with the associated data-sets. .
Some of our findings include:

1. Ten countries in Africa account for 60% of all Internet disruptions experienced in the last five years.
2. All countries that have had an Internet disruption have had the current ruling party being in power for 18.9 years on average.
3. Countries with less than 20% Internet Penetration rates are more likely to disrupt the Internet during protests than those with higher rates.
4. Liberal countries are less prone to Internet disruptions, especially where sufficient oversight exists over the executive arm of Government.
5. Detection and attribution of Internet disruptions is improving but regional disruptions remain a daunting task. 

We were also interested in estimating economic impact of intentional disruptions in African countries. The report shows that by incorporating ‘shadow economy’ in assessing impact of Internet disruptions, there is an average of as high as 30% jump in economic costs from previous estimate models. The ‘shadow economy’ is understood here as economic activities and the income derived that circumvents or otherwise avoids government regulation, taxation or observation (Schneider 2013). This includes what we are calling the ‘WhatsApp Economy‘, that involves individuals or small businesses using messengers (especially WhatsApp and Telegram) and social media platforms (especially Facebook, Instagram, and Twitter) to market their wares or services, aided by mobile money and boda boda (motorbike couriers) to complete transactions without any registered business or additional tax responsibilities.
The first section conducts an audit of how Internet disruptions have been defined, detected, attributed, costed and responded to. Section two looks into how to quantify effects of Internet disruptions in Africa. Section three presents the findings from the quantification exercise and section four discusses some cases from the findings and section five presents research and policy recommendations.
Download the report here.