At a time when many people around the globe are dependent on digital networks to work from home, stay connected with family and friends and get the latest information on the worst global health crisis in generations, the thought of not being able to access the internet is daunting. But in “normal” times, network disruptions and large-scale shutdowns are not uncommon. The phenomenon has actually been on the rise in recent years, with approximately 213 shutdownsin 2019 globally, up from 196 in 2018 and 109 in 2017. These incidents come at a considerable cost: for a one-day complete network shutdown in the United States, the NetBlocks Cost of Shutdown Tool (COST) calculates a staggering $7,286,617,109. According to the same tool, in fragile and conflict situations, a shutdown would cost an estimated $209,412,474 per day. Network shutdowns are therefore overwhelmingly political rather than economic. Jan Rydzak, research analyst at Ranking Digital Rights and a recent UN Data Innovation Specialist, explains why such shutdowns occur, what the political implications are and their impact on peace and security.
When do network shutdowns typically occur? And what are some of the common justifications for network shutdown?
Jan Rydzak: There have been approximately between 600 and 700 shutdowns since 2011, when the first most notable shutdowns were recorded, across more than 50 countries. By and large, shutdowns take place around perceived or real security situations. Governments typically use the justification of safety concerns, including around real or potential violence. It’s convenient that the burden of evidence is not very great on them as it stands. There’s no mechanism in any country that would compel governments to actually prove that there is a legitimate public safety concern involved. So, protests or potential protests make up between 80 and 90% of the cases of network shutdowns that we see around the globe. Now, that often overlaps with particular political events, and most notably, many shutdowns have occurred around elections, especially in Sub-Saharan Africa. Governments, again, justify them with the explanation that elections are a unique circumstance in which disinformation is a particularly salient concern and may actually have some bearing on the immediate political future of the country.
We see a fair number of shutdowns that take place around mass events that are not necessarily political, for example religious processions, which have coincided with shutdowns especially in India and Pakistan. There are also increasingly arbitrary events. There’s been at least one shutdown around a wrestling match, and one shutdown in the Philippines around a beauty contest. This is a slow advancement towards more and more spurious justifications that increasingly move away from legitimate public safety concerns.
And then, possibly the most surprising category: Several countries, notably Ethiopia, Syria, Algeria and a couple of others, have actually shut down the internet – the entire internet infrastructure – around both professional and school exams that have taken place across the country. It’s hard to argue that this is the least disruptive way to deal with the concern of cheating on national high school exams – hardly the most appropriate measure to use against infractions like that.
The main reasons cited for shutting down the internet are safety concerns, but as your study around network shutdowns and protest movements in India shows, they usually don’t achieve their stated goals. Why is that?
They’re not effective in safeguarding public safety because they’re not fit for purpose. Shutdowns are, and this is a point that has to be driven home, extremely large-scale events that affect hundreds of thousands to millions of people. And so, alongside any potential favorable effects for public safety, there will be a host of additional ramifications that will work against that original purpose.
One of the dynamics that are at play here is that shutdowns turn situations that are often already tense and chaotic into even greater chaos. They exacerbate the existing levels of chaos, inject even more instability into unstable situations. In some cases that we’ve noted (most notably the research on India), they have been found to not curtail violent protests but to actually lead to a tactical shift towards less orderly, more chaotic forms of protest – and often, unfortunately, to more and more violence. In India, analyzing about 23,000 conflict events in 2016, I found that shutdowns are associated with just such a tactical shift toward violent unrest – not just on the first day, but for several days afterwards. In the information vacuum that ensues, it takes just as long for violent incidents to die down as when connectivity is not severed, but they erupt more frequently. At the same time, shutdowns appear to have very inconsistent effects on peaceful demonstrations, to the point where they are no more successful at discouraging such gatherings than random chance. And it is worth remembering that peaceful, public dissent is explicitly protected by international human rights law.
The last piece of the puzzle, which applies to more than just individual prominent cases, is that some of the platforms that are shut down in such situations are also platforms that are often used to coordinate and organize, such as Facebook, Twitter, or WhatsApp. And typically, a protest that is organized – for example, through these digital tools – tends to be more structured and, by extension, more peaceful. One can argue that, if a government actually chooses to shut down these platforms, the most they can expect is that the blackout may disrupt peaceful protests, but certainly will not mitigate, quell, or preempt escalations of violence, which is supposedly the effect that most governments aim to achieve.
There are quite a few examples of this. The most striking recent one was the case of Sudan. It’s also an interesting case because it allows us to illuminate a different aspect of the same phenomenon. So, in Sudan, during the turmoil that occurred between the eruption of large-scale protests against President Omar al-Bashir in December 2018 and the dissolution of the Transitional Military Council’s junta government in mid-2019, the two governments that ruled Sudan in this period imposed three separate shutdowns. One of them was a shutdown that ran from December to February for three full months and severed access to most social media. Then there was another, more ephemeral shutdown that the Bashir regime imposed in April. And finally, the military junta that took power after the fall of the Bashir regime imposed a full internet shutdown at the beginning of June. In all of those cases, pre-existing protests were predominantly peaceful – and after the shutdowns began, in all three cases, there was actually an increase in the concentration and number of peaceful protests. There can be no doubt about the fact that both governments were trying to get people off the streets, trying to discourage protests, but the protests continued unabated – in part because there were already pre-existing ways to coordinate that were not necessarily reliant on these digital networks. So, the backfire effect was a little different in this case, in the sense that there was no spike in chaotic, disorganized violence, but there were increases, and actually waves of increases, in peaceful protests which ultimately ended up leading to the collapse of the Bashir regime.
What is the impact of network shutdowns on peace and security other than just shifting protests to other forms or sometimes more violent forms of protests? And what are some of the trends you’ve seen around the globe?
Well, there are several aspects to this. If shutdowns are meant as a measure to defuse violence by organized or armed non-state actors, then there’s evidence to show that they’re not effective in that either. The first notable example of this ineffectiveness that I have seen occurred in Nigeria. In the course of the fight against Boko Haram in 2013, the government imposed a very large-scale cellphone shutdown in three regions in the northeast of the country. There were several interesting aftereffects of this shutdown, according to research published a couple of years later. The shutdown led to a strategic shift within Boko Haram to a more closed, centralized insurgency, and they shifted their geographical base of operations to the Sambisa Forest. It backfired in the sense that that the terrorist group started using circumvention tactics and adapting to this new landscape of communication without technology. This is something that we have seen time and time again.
The second effect of this shutdown was that it may have been linked to loss of life that could have otherwise been avoided, namely the Benisheik massacre in Borno State, Nigeria. News reports showed that approximately 90% of the almost 200 people that were killed in that event may have been traveling in order to regain access to connectivity. Similarly, in Cameroon, a prolonged shutdown in 2017 forced many people to undertake perilous journeys across unstable territory to regain access in makeshift border basis; some news outlets later dubbed them “internet refugees.” Disrupting access to these communication networks – not just to individual types of messaging but to the entire channel – means that what you end up doing is depriving people of their sources of information, including information that may contribute to their own physical security. And that has a direct effect on people’s most fundamental human right: the right to life.
Another security dimension of network shutdowns is that in several cases, we’ve seen that shutdowns are used as kind of an invisibility cloak for human rights abuses. In Syria, for example, shutdowns were timed to coincide with military operations in areas that were predominantly controlled by the opposition. So, it was a blackout that could be likened to a military tactic or communications siege. Similarly, in the case of Sudan: The massacre that occurred on 3 June 2019 was immediately followed by a complete shutdown of all internet connectivity across the country aimed at blocking the flow of information that would reveal the extent of the violence that was committed by the regime. But shutdowns don’t actually stop the flow of information – they delay it. Videos of soldiers shooting at unarmed protesters in Khartoum existed regardless of the fact that the internet was cut off for the space of a month – and, of course, those videos surfaced online after connectivity was restored. That’s another way in which shutdowns are ineffective.
What is the effect of network shutdowns when used during elections?
Governments have imposed shutdowns around elections for years, with and without accompanying violence, and almost always under the pretense of stopping disinformation and the dissemination of false results. Some of the more recent examples include Benin, Mauritania, Bangladesh, and Togo. While there is not a whole lot of research on this, there seems to be a link between state ownership of internet service providers (ISPs) and the likelihood of shutdowns amid electoral violence. Although there’s no consistent evidence of shutdowns always favoring the incumbent, it would make sense that if you block this multilateral channel of communication, you’re going to end up underlining the existing channels that are in most cases controlled by the government. In the context of elections that obviously gives the incumbent an edge, because channels that are otherwise effectively used by opposition candidates (such as the Internet or specific apps) are simply unavailable and people resort to information channels that are often already geared towards supporting the government. This is one of many ways in which shutdowns can skew the playing field, aside from simply cutting off access to information about candidates and their positions.
Are there any “good”, or justifiable, reasons to shut down networks?
I’m not alone in the contention that shutdowns are not justifiable in any circumstances. There have been at least two UN Special Rapporteurs – the Special Rapporteur on Freedom of Opinion and Expression and the Special Rapporteur on Freedom of Assembly and Association – who have called multiple times for bringing an end to shutdowns. In their extensive consultations, neither of them could not find any evidence of justifiable circumstances in which shutdowns could take place. In situations of public safety and national security, we haven’t had a single case where governments actually showed compelling evidence that this is the right tactic to use. And if they don’t meet that test, including in situations of potential violence that come closest to legitimacy, then there is likely no circumstance under which shutdowns are acceptable.
There’s little disclosure from governments about the effects of shutdowns. There are two ways to interpret this: One of them is that they simply don’t know what the effects are, and the other is that they know, but they won’t tell. They might know that this tactic is not actually effective and that the goal might actually be to demonstrate power. Shutdowns are a show of force that displays what the government can do, and they are a kind of restriction that, unfortunately, can be relatively easy to execute and immediately affects large swathes of the population. It’s a low-cost way to control all communication at once. And especially with rising connectivity, the impression it makes is going to reverberate across the country.
Are there any gender dimensions to network shutdowns?
Yes, and some of those dimensions connect to larger trends that affect vulnerable communities and groups in general. Most notably, there’s research on the effects of shutdowns on the lives of women in a state in India, in Manipur. The report’s findings suggest that women are disproportionately affected - especially in terms of personal and economic security. This drives home the point that shutdowns strip people of access to critical tools and mechanisms that low-cost connectivity enables - like mobile banking networks, which are the primary conduit of remittances in many countries.
Have you seen any impact of the current pandemic on the decision to shut down networks?
As of 24 April 2020, there are still, as far as I am aware, two major shutdowns that are ongoing. One of them is in Myanmar and the other is in India. The Myanmar shutdown is affecting the same townships and the same states, Rakhine and Chin, that it has been affecting for the last eight months. And the shutdown in India is concentrated in Kashmir, where the most that the local population can count on is 2G connectivity – and even that gets shut down during violent skirmishes. In those two cases, we have governments – in one case a national government and in the other case a state government supported very strongly by the national government – that are intransigent in terms of their insistence on shutdowns. There’s been one glimmer of hope: the shutdown that was ongoing in the Oromia region in Ethiopia has been called off. I think more than ever, both the telecommunications companies, civil society and activists are calling for governments to refrain from conducting shutdowns in such a unique time of universal quarantine. But the decision to call off a long-standing shutdown is tricky, because it would involve governments having to admit that these networks are actually useful for the transmission of helpful information, - that shutdowns affect not just disinformation and disinformation flows, but more than anything, credible and accurate information, which I think is necessary now more than ever.
Disclaimer: The views in this article do not necessarily reflect positions of the United Nations.