This isn't really true. Take for example the Harvard Study on Chinese Censorship "Reverse-engineering censorship in China: Randomized experimentation and participant observation" by King, Gary, Pan, and Roberts (http://gking.harvard.edu/files/gking/files/experiment_0.pdf).
Scathing criticism is regularly posted about the Chinese government and is not subject to any censorship. There are a few key areas that are censored: notably relations with Tibet and historical political uprisings.
> Scathing criticism is regularly posted about the Chinese government
Did you actually read the study? That's not what they studied. They studied whether postings were held for review more on social media sites over standard media. And they weren't held more. What they found was that the CCP on social media takes a "post first, censor later" approach. That doesn't mean they don't censor social media. On social media, posts are "Reviewed".
If you scroll down to the bottom of the study, they even give a listing of words that usually trigger a "review" flag.
From your linked study: "In total, 66 of the 100 sites in our sample (automatically) review at least some social media submissions, and
40% of all of our individual social media submissions from our 100 sites (and 52% of submissions from sites that review at least sometimes)
are put into review. Of those submissions that go into review, 63% never appear on the web."
Did you actually read the paper? You quote from the results section. Did you go directly there?
Here's the opening paragraph. You'll find plenty in the paper to support it, although this is directly supported with a citation from their previous paper (http://gking.harvard.edu/files/gking/files/censored.pdf): "We begin with the theoretical context. The largest previous study of the purpose of Chinese censorship (2) distinguished between the “state critique” and “collective action potential” theories of censorship and found that, with few exceptions, the first was wrong and the second was right: Criticisms of the government in social media (even vitriolic ones) are not censored, whereas any attempt to physically move people in ways not sanctioned by the government is censored. Even posts that praise the government are censored if they pertain to real-world collective action events (2)."
Critique of the state is not censored. The paper goes on to show that even discussion about having a multiparty American-style Constitutional system is not subject to censorship.
There are very select things that are censored, and these things that are censored (calls for direct collective action and specific topics related to these) see both:
1.) foreign propaganda attempts, which is a strong reason for the Chinese government to be interested in disruption (this is also true in America)
2.) parity with movement disquieting behavior of Western states (the US disquiets movements through mechanisms other than direct censorship including direct organizational infiltration, Fusion Centers, IRS harassment, JTRIG type 'warnings', character assassination, bribery, etc)
Scathing criticism is regularly posted about the Chinese government and is not subject to any censorship. There are a few key areas that are censored: notably relations with Tibet and historical political uprisings.