This (last for now) literature review will address the concept of a “complete network shut-down”, which the article “When Do States Disconnect Their Digital Networks? Regime Responses to the Political Uses of Social Media” brings to the reader’s attention. It will be considered here, firstly, what kind of situation does this term apply to; secondly, it will be assessed whether a “complete network shut-down” can be understood as a part of the concept of “filtering” proposed by the authors of “The Dictators Digital Toolkit: Explaining Variation in Internet Filtering in Authoritarian Regimes”, and, finally, it would be considered why a complete network shut-down is or is not a preferred strategy of authoritarian regimes when dealing with dangerous content.
To begin with, one has to note that it is possible that the word “complete” is slightly misleading. It does not refer, as the examples of Haiti in 1999, China in 2006, Egypt in 2011 in Regime Responses to the Political Uses of Social Media’s demonstrate, to a state stopping any and all access to the Internet within its sovereign territory. Instead, it is more of a controlled blackout, where a certain region or a certain group of people are temporarily disconnected from a network. Their purpose, so it seems, is to retain control of a situation by restricting the spreading of particular information within a country, as well as overseas. Furthermore, drawing on the section pertaining to the BlackBerry scandal in the UAE, one could suggest that this attempt of severing ties between a group of users or servers with access to this information and the rest of the Internet could often constitute a desperate attempt of a government to save face and “keep a lid on” compromising data or information. When seen from this perspective, this measure is both a preventive one, as the government tries to act as promptly as it can to nip the information loss in the bud, and, potentially, also a last-minute one, filled with desperation of a drowning person who is grasping at a straw.
An interesting question to ask oneself is if shutting down a part of the network can be understood as a “filtering tool”, at least in the sense Hellmeier’s article uses that expression. When approached purely linguistically, a “filtering” strategy seems to indicate a mechanism much finer than the brute force of a “shut-down”. In relation to the strategies of governmental control of traditional media as portrayed by Regime Responses to the Political Uses of Social Media, an act similar to a network shut-down would be “destroying newspaper supplies and seizing radio and television stations”, whereas “filtering” would be more akin to indirect economic pressure of media outlets, censorship of newspapers and TV channels, and journalists’ own self-censorship. Drawing this parallel underlines a clear difference in strategy and degree of impact between the two. However, there is an argument to be made in favor of including “complete” network shut-downs into the term “filtering”, too. The Dictator’s Digital Toolkit mentions “server takedown” as a filtering mechanism; this strategy perfectly captures the situation reported in Howard’s, Hussain’s and Agarwal’s article, namely the call of the German prosecution officers in 1995 to restrict access to a specific resource: in that case, one had identified potentially harmful content and moved through the proper channels, namely the judicature, to obtain a permission to take that content down. Also the desired effect of network shut-downs and filtering (in its broadest sense) is very similar: in both cases one tries to obstruct the process of certain information becoming known to the wider public, just in two different ways – either concealing it by basically making it invisible to local search engines, as what The Dictator’s Digital Toolkit seems to suggest with “dropping data packages according to destination, origin or banned keyword”, or, as is the case with shut-downs, preventing its sharing by “pulling the plug”. Moreover, an attempt to distinguish the two – filtering and shut-downs – based on a criteria of legality (or, to put it better, of whether an action is performed in accordance with the rule of law, this is to say “within a legal system that has some morality built into it”, as even the most despotic regime is capable of creating law) does not bear any fruits: just as both “softer” censorship can take place in democracies, fragile states and autocracies, shut-downs are fairly common in both regimes (see Regime Responses to the Political Uses of Social Media, Table 1). That all being said, though functionally the two may have a similar purpose, Howard’s, Hussain’s and Agarwal’s distinction between complete network shut-downs and “targeting individual websites or their users” seems to firmly place network shut-downs in a standalone category of activity pertaining to content regulation.
According to Howard’s, Hussain’s and Agarwal’s research, esp. Table 1, bringing the complete network down is not the governments’ preferred method of choice of dealing with dodgy content, and it is easy to see why. Firstly, as portrayed by the events of the uprising in Egypt in 2011, a shut-down is not always efficient: just as a mouse can squeeze through the tiniest of holes, all information needs is a single transmission channel that is still up and running, and, when taking the grassroots character of a popular uprising into account, it is only logical to assume that the protesters will try their best to find a way to get a message of what is going on locally out to the world, and will most likely find it. Furthermore, as was demonstrated by protests in Egypt, an authoritarian regime attempting a network shut-down needs to ensure the absolute loyalty of the very officials executing those orders, and command them to act promptly and efficiently, which is not always the case. Therefore, unless supported by some kind of repressive measures of an effective far-reaching state, a shut-down perhaps more often than not will not be worth the resources spent on it. Another argument as to why states seem to favor softer, filtering measures is the question of legitimacy: since it is crucial for an authoritarian state to mask its censorship attempts as “protection of public good” or “national security”, a network shut-down is too sloppy and risks damaging the state’s reputation. Finally, a harsh measure like that is bound to attract international attention, and authoritarian states should tend to avoid international scandals, because they lead to tribunals and investigations, which such a state cannot permit.
As a concluding remark, another reason for why authoritarian states seem to favor less drastic web-content manipulation measures, may perhaps lie in the character of the Internet itself. In his article on why securitization of Internet is a hard and a never-ending task, Schulze notes that the way one is currently attempting to make the Internet secure is by storing large amounts of data and meticulously going through them. (Unsicherheit hinter dem Bildschrim, Shulze in Innere Sicherheit nach 9/11, 2016 [DE]) In doing so, one tries to “separate” parts of an ever-expanding medium and by making those parts secure, make the medium more secure as well. However, that is Sisyphus’s work that never ends, because there are always ways to make “secured” content dangerous just by the virtue of interconnectedness of the Internet. In the same vein, just as securitization by limitation does not work, taking out a part of the Internet and imposing a blackout on it can hardly be done, as one cannot completely sever the connection between it and the rest of the global network. This is why less drastic measures that do not strive to separate the “flagged” content, but rather mask it, seem to be more effective and in line with what the Internet is. This may also make authoritarian states inclined to use censorship as opposed to bringing down the complete network.
Cover image source: https://www.zambianobserver.com/southern-province-in-internet-network-shut-down/
Just how “complete” a “complete network shut-down” really is? To what extent is filtering the Internet a more desirable strategy for an authoritarian state over a complete network shut-downs?
It's impossible in my country but I hope it'll happen at least once.