The Chinese Trolls Who Pump Out 488 Million Fake Social Media Posts

New Research Exposes 50-Centers, a ‘Massive Secretive Operation’ to Fill China’s Internet with Propaganda

They are the most hated group in Chinese cyberspace. They are, to hear their ideological opponents tell it, “fiercely ignorant,” keen to “insert themselves in everything,” and preen as if they were “spokesmen for the country.” Westerners bemoan their propensity to beat the drum of nationalism, bombard Chinese liberals with personal attacks, and pollute online dialogue with wave after wave of strident propaganda. While their ranks have been unknown and their precise inner workings uncertain, at least everyone agrees on their name: wumao, or 50-centers (also called the 50 Cent Party), slang for the 50 Chinese cents they allegedly receive for each social media post. Now, a new report suggests it’s time to re-imagine who these people are and how they operate.

A May 17 paper written by professors at Harvard, Stanford, and the University of California, San Diego provides the most detailed and ambitious description of China’s 50-centers available to date. It confirms the existence of a “massive secret operation” in China pumping out an estimated 488 million fabricated social media posts per year, part of an effort to “regularly distract the public and change the subject” from any policy-related issues that threaten to anger citizens enough to turn them out onto the streets. But the research finds no evidence these 50-centers are, in fact, paid 50 cents, nor does it find they engage in direct and angry argument with their opponents. Instead, they are mostly bureaucrats already on the public payroll, responding to government directives at a time of heightened tension to flood social media with pro-government cheerleading.

In countless online chat rooms, bulletin boards, and Weibo threads, Chinese social media roils with the same ideological debates that also increasingly consume Chinese academics and elites.

“The content of [50-center] posts was completely different than what had been assumed by academics, journalists, activists, and participants in social media,” Jennifer Pan, an assistant professor at Stanford and one of the report’s authors, told Foreign Policy. “They—and we before we did this study—turned out to be utterly wrong” about how pro-government shills actually operate.

Understanding the behavior of pro-government netizens is important, given the stakes. In the past two and a half years, the Chinese government has used a combination of muscle and guile to cow online opinion leaders into submission, muzzling social media as a political force, and leaching public dialogue of much of its independence. But beneath the peppy, pablum-filled surface that has resulted, Chinese social media remains a contested space. In countless online chat rooms, bulletin boards, and Weibo threads, Chinese social media roils with the same ideological debates that also increasingly consume Chinese academics and elites.

Broadly speaking, the clash pits so-called leftists—that is, conservatives and neo-Confucianists who marry stout Chinese nationalism, a yearning for reconstructed socialism, and the quest for a reversion to hierarchy and filial piety—against rightists, or reformists, who continue to espouse what a Westerner would recognize as universal values, such as civil and human rights, government transparency, and democracy and constitutionalism. It’s more common for the two camps to exchange barbs than ideas. The leftists label the rightists sellouts, turncoats, and “public intellectuals,” the latter delivered with an implicit sneer. The rightists often call the leftists “50-centers,” regardless of who really pays their bills.

Given the infighting, it’s not hard to picture a shadowy coterie of young, angry, and irremediably argumentative 50-centers pitted against the nation’s liberals. Actual 50-centers, it turns out, are also far less likely to trade arguments or insults with their interlocutors than they are to stream peppy drivel into major discussions at just the right time. Of the posts the researchers analyzed, 80 percent were labeled “cheerleading,” and 13 percent “non-argumentative praise or suggestions.” These include such barn-burners as, “We all have to work harder, to rely on ourselves, to take the initiative to move forward” and, “We hope the central government provides us with even more support.” There’s little to offer such blather beyond a shrug or a grunt—that, of course, is precisely the point.

Although the number of fabricated posts is impressive, it’s also small compared with the heaving corpus of approximately 80 billion posts generated on China’s hyperactive social media each year.

Although the number of fabricated posts is impressive, it’s also small compared with the heaving corpus of approximately 80 billion posts generated on China’s hyperactive social media each year. And 50-centers spend about half their energy posting on the friendly terrain of government-run websites. That means that only one out of every 178 posts on commercial Chinese social media actually comes from a 50-center. To maximize influence, the commentary mostly emerges at times of particularly intense online discussion, when the volume of chatter spikes—and when, the report’s authors argue, the possibility of online protest emerging into the real world is highest. (Disappointingly, researchers do not attempt to estimate the total ranks of 50-centers.)

Media

05.31.12

Godwin’s Law with Chinese Characteristics

Hu Yong
This winter writer-blogger-race car-driver Han Han found himself facing charges of plagiarism from celebrated fraud-buster Fang Zhouzi. Both Han and Fang have huge followings among China’s microbloggers. And their personal disagreement soon...

The path to unmasking the 50-cent group began with a December 2014 leak of emails emanating from the Internet Information Office of Zhanggong district in Ganzhou, a small city in the southeastern province of Jiangxi. Researchers sought to identify how many of those named in the leak were actually 50-centers. Smoking out these notorious pro-government trolls didn’t require too much derring-do; researchers simply asked them by creating pseudonymous social media accounts, then direct-messaging those named in the leaked documents with this message: “I saw your comment, it’s really inspiring. I want to ask, do you have any public opinion guidance management, or online commenting experience?” (The anodyne term “public opinion guidance management” is widely recognized as the government’s code word for 50-centers.) Many, perhaps flattered by the approach, were happy to respond by admitting what they did. Pan said she was “not particularly surprised” the subjects were so forthcoming. “If you participate in online sentiment guidance, you might see yourself as someone who helps improve the general tenor of online discussions—this would not be something to be embarrassed about or ashamed of,” she said.

Beyond the eye-popping numbers, this report may prove most useful in outlining the limits of 50-center scope and influence. The notion that a massive, paid army of truculent pro-government netizens is largely to blame for China’s impoverished public dialogue is Orwellian, yet strangely comforting. If most of the pro-government invective calling reformists traitors to the motherland is driven by government dictat, that makes it possible a less illiberal regime could turn off the spigot of venom, allowing more recognizably Western views to thrive. But this report implies that those espousing nationalist nastiness aren’t paid shills after all. They mean precisely what they say.