New Research: Internet Censorship To Stop Protests... Actually Increases Protests
from the who-didn't-see-that-coming? dept
The key part of the research is that it models the behavior of protesters in various cases, using proven simulation techniques. It's really quite fascinating to see how they model "civil violence." While simplified, it definitely can be useful for modeling how people react in certain situations, and it appears to have a decent track record.
The agent's behaviour is influenced by several variables, the first one being his/her personal level of political dissatisfaction ("grievance", indicated by lighter or darker shades of green colour in figure 1). This can lead the agent to abandon his/her state of quiet and become an active protester (red coloured circles in figure 1). However, the decision to act out -- whether it is to go on a looting spree or to engage in violent demonstrations -- is conditioned by the agent's social surroundings ("neighbourhood" in the model's language). Does s/he detect the presence of police (blue triangles in figure 1) in the surroundings? If the answer to this question is no, s/he will act out. If the answer is yes, another question is asked: is this police presence counterbalanced by a sufficient number of actively protesting citizens? If the answer to this second question is yes, then the agent acts out. Sometimes, in an utterly random way, one of the active citizens gets caught by the police and is sent to jail for a given period of time (black circles in figure 1). Again, the apparent simplicity of this rule, is sadly consistent with the many episodes of arbitrary and "swift" justice triggered by the UK government adoption of a hard line in dealing with riot repression.From there, the researchers, Antonio Casilli and Paola Tubaro, model in internet censorship to see what impact it has. What they find is the greater the censorship, the greater the violence. The least amount of violence occurs when there is no censorship. While the model could certainly be improved upon, it is useful to see that at least the first attempt at modeling such behavior seems to indicate what we've been saying for a while: censorship isn't likely to help and likely makes things worse.