Professor Gary King at Harvard, together with Assistant Professor’s Jennifer Pan from Stanford and Margaret E. Roberts from the University of California have done several remarkable things in the paper highlighted today.
Not only have they produced the first in depth study of how the Chinese authorities attempt to manipulate views on social media they’ve also managed to put numbers on just how big the effort is. Moreover, in the process they deliver strong push back on a number of widely held misconceptions about this activity held by the media, academics and ordinary Chinese netizens.
First, how big? They estimate of a total of c. 80bn posts on social media in China annually around 448mn are bogus. Just over half are posted on government sites and the rest on services like Sina Weibo. Put another way, one in 178 posts on Chinese social media is a plant of some sort.
Just how wrong is conventional wisdom on the activities of the posters or the so called ‘Fifty Cent Party’ (FCP) (it’s believed the government pays goons this sum for every post)? The academics address four common misconceptions.
1) The FCP stays away from direct argument or debate with other netizens. It’s widely believed that if you get into a tussle with somebody online about a sensitive matter an FCP-er is on the other end of the discussion. The researchers found no evidence of this. In fact the FCP spend most of their time trying to steer conversations away from difficult issues by creating diversionary-topic spikes on other subjects.
2) The FCP does not attempt to defend either the government or leadership from criticism. It’s been acknowledged for some time the government find criticism, particularly if directed at potentially corrupt local administrators, helpful. The study confirms this to be the case and can find no evidence of engagement by the FCP with other netizens on these issues.
3) The FCP are discovered to be not journeyman for hire but in fact government employees. The study uses source material from a leak by a local propaganda division from a small(ish) city in Jiangxi which has given them reliable date to extract a great deal of color on individual FCP-ers. The leaked material had many instances of government officials reporting back to high-ups about quotas being filled.
4) Having established the FCP is not a legion of goons working out of a shed in Liaoning the researchers discover the very name of the movement is fundamentally incorrect. There is no evidence of fabricated content providers being paid 50-cents a pop or receiving any specific pay at all for this activity. Their efforts may of course form part of a general work assessment but pay-per-post? No.
For such a content rich paper its a refreshingly short and well written without the scary deep dive into methodology par for the course in much of this sort of work. You can access it in full via the following link Fabricating Social Media.
Happy Sunday