The shutdown affected the productivity of entrepreneurs and businesses in Benin. Young entrepreneurs who needed the Internet to use platforms like Slack or Whatsapp, or perform tasks like updating their servers, were left paralysed by the blackout
Access to social media, including WhatsApp, Facebook, Twitter, Telegram, and Instagram was reportedly blocked at midnight of April 28, as Benin prepared to hold parliamentary elections without opposition parties. A full Internet blackout by Internet service providers, including Spacetel, MTN and Moov, followed at around 10:00 in the morning.
The shutdown has not been publicly acknowledged by the government, but local reports were corroborated by network measurement data from Open Observatory of Network Interference (OONI), the Center for Applied Internet Data Analysis (CAIDA) and RIPE Atlas – organizations that track interruptions in connectivity worldwide.
Virtual private networks including OpenVPN, Hola and Tunnelbear, which are used to circumvent Web blocking, were also inaccessible in the country.
The following chart aggregates OONI measurement coverage from the testing of WhatsApp in Benin in April 2019. While the overall measurement coverage was very low (limited to measurements collected between 27th to 29th April 2019), we observe that most measurements collected on 27th and 28th April 2019 presented anomalies, signalling WhatsApp blocking.
Chart: OONI Probe testing of Whatsapp in Benin in April 2019 (source: OONI MAT).
Chart: Per ASN breakdown of OONI Probe testing of WhatsApp in Benin in April 2019 (source: OONI MAT).
We observe that while WhatsApp was found accessible on other tested networks, it only presented signs of blocking on Spacetel Benin (AS37424).
Similarly, OONI data shows signs of Telegram blocking in Benin on 27th and 28th April 2019, with most anomalies only occurring on Spacetel Benin (AS37424).
Chart: OONI Probe testing of Telegram in Benin in April 2019 (source: OONI MAT).
On the other hand, the testing of Facebook Messenger presents a slightly different picture, as we mainly observed anomalies on 29th April 2019 on Jeny SAS (AS328098).
Chart: Per ASN breakdown of Facebook Messenger measurement coverage in Benin in April 2019 (source: OONI MAT).
Meanwhile, the testing of www.facebook.com only presented anomalies on 28th April 2019 (election day).
Chart: OONI Probe testing of www.facebook.com in Benin in April 2019 (source: OONI MAT).
Similarly, OONI data shows that the testing of Instagram only presented anomalies on election day (28th April 2019).
Chart: OONI Probe testing of Instagram in Benin in April 2019 (source: OONI MAT).
Also on election day (28th April 2019), OONI data shows that the testing of Snapchat presented signs of blocking.
Chart: OONI Probe testing of Snapchat in Benin in April 2019 (source: OONI MAT).
In addition to these social media blocks, OONI data suggests that access to certain circumvention tool sites may have been blocked in Benin amid the elections.
Chart: OONI Probe testing of circumvention tool websites in Benin in April 2019 (source: OONI MAT).
The above graph suggests that access to the websites of OpenVPN and TigerVPN may have been interfered with during the elections, but it’s worth noting that other circumvention tool websites (such as torproject.org, psiphon.ca, and purevpn.com) were accessible on tested networks the next day (29th April 2019).
Overall, the limited measurement coverage presents an important limitation to the findings. Yet, the fact that several social media websites presented anomalies during the same dates provides a signal of blocking, potentially corroborating reports from the ground.