(PHOTO: Marcus Bleasdale VII)
Show Hide image

China’s Paid Trolls: Meet the 50-Cent Party

The Chinese government hires people to distort or deflect conversations on the web. Ai Weiwei persuades an “online commentator” to tell all.

In February 2011, Ai Weiwei tweeted that he would like to conduct an interview with an “online commentator”. Commentators are hired by the Chinese government or the Communist Party of China to post comments favourable towards party policies and to shape public opinion on internet message boards and forums. The commentators are known as the 50-Cent Party, as they are said to be paid 50 cents for every post that steers a discussion away from anti-party content or that advances the Communist Party line.

Below is the transcript of Ai’s interview with an online commentator. As requested, an iPad was given as compensation for the interview. To protect the interviewee, relevant personal information has been concealed in this script.

Question: What’s your name, age, city of residence and online username?

Answer: I cannot make my name public. I’m 26. I have too many usernames. If I want to use one, I just register it. I won’t mention them here.

What do you call the work you do now?

It doesn’t matter what you call it: online commentator, public opinion guide, or even “the 50-Cent Party” that everyone’s heard of.

What is your level of education and work experience? How did you begin the work of guiding public opinion?

I graduated from university and studied media. I once worked for a TV channel, then in online media. I’ve always been in the news media industry, for four or five years now.Over a year ago, a friend asked me if I wanted to be an online commentator, to earn some extra money. I said I’d give it a try. Later, I discovered it was very easy.

When and from where will you receive directives for work?

Almost every morning at 9am I receive an email from my superiors – the internet publicity office of the local government – telling me about the news we’re to comment on for the day. Sometimes it specifies the website to comment on, but most of the time it’s not limited to certain websites: you just find relevant news and comment on it.

Can you describe your work in detail?

The process has three steps – receive task, search for topic, post comments to guide public opinion. Receiving a task mainly involves ensuring you open your email box every day. Usually after an event has happened, or even before the news has come out, we’ll receive an email telling us what the event is, then instructions on which direction to guide the netizens’ thoughts, to blur their focus, or to fan their enthusiasm for certain ideas. After we’ve found the relevant articles or news on a website, according to the overall direction given by our superiors we start to write articles, post or reply to comments. This requires a lot of skill. You can’t write in a very official manner, you must conceal your identity, write articles in many dif­ferent styles, sometimes even have a dialogue with yourself, argue, debate. In sum, you want to create illusions to attract the attention and comments of netizens.

In a forum, there are three roles for you to play: the leader, the follower, the onlooker or unsuspecting member of the public. The leader is the relatively authoritative speaker, who usually appears after a controversy and speaks with powerful evidence. The public usually finds such users very convincing. There are two opposing groups of followers. The role they play is to continuously debate, argue, or even swear on the forum. This will attract attention from observers. At the end of the argument, the leader appears, brings out some powerful evidence, makes public opinion align with him and the objective is achieved. The third type is the onlookers, the netizens. They are our true target “clients”. We influence the third group mainly through role-playing between the other two kinds of identity. You could say we’re like directors, influencing the audience through our own writing, directing and acting. Sometimes I feel like I have a split personality.

Regarding the three roles that you play, is that a common tactic? Or are there other ways?

There are too many ways. It’s kind of psychological. Netizens nowadays are more thoughtful than before. We have many ways. You can make a bad thing sound even worse, make an elaborate account, and make people think it’s nonsense when they see it. In fact, it’s like two negatives make a positive. When it’s reached a certain degree of mediocrity, they’ll think it might not be all that bad.

What is the guiding principle of your work?

The principle is to understand the guiding thought of superiors, the direction of public opinion desired, then to start your own work.

Can you reveal the content of a “task” email?

For example, “Don’t spread rumours, don’t believe in rumours”, or “Influence public understanding of X event”, “Promote the correct direction of public opinion on XXXX”, “Explain and clarify XX event; avoid the appearance of untrue or illegal remarks”, “For the detrimental social effect created by the recent XX event, focus on guiding the thoughts of netizens in the correct direction of XXXX”.

What are the categories of information that you usually receive?

They are mainly local events. They cover over 60 to 70 per cent of local instructions – for example, people who are filing complaints or petitioning.

For countrywide events, such as the Jasmine Revolution [the pro-democracy protests that took place across the country in 2011], do you get involved?

For popular online events like the Jasmine Revolution, we have never received a related task. I also thought it was quite strange. Perhaps we aren’t senior enough.

Can you tell us the content of the commentary you usually write?

The netizens are used to seeing unskilled comments that simply say the government is great or so and so is a traitor. They know what is behind it at a glance. The principle I observe is: don’t directly praise the government or criticise negative news. Moreover, the tone of speech, identity and stance of speech must look as if it’s an unsuspecting member of public; only then can it resonate with netizens. To sum up, you want to guide netizens obliquely and let them change their focus without realising it.

Can you go off the topic?

Of course you can go off the topic. When transferring the attention of netizens and

blurring the public focus, going off the topic is very effective. For example, during the census, everyone will be talking about its truthfulness or necessity; then I’ll post jokes that appeared in the census. Or, in other instances, I would publish adverts to take up space on political news reports.

Can you tell us a specific, typical process of “guiding public opinion”?

For example, each time the oil price is about to go up, we’ll receive a notification to “stabilise the emotions of netizens and divert public attention”. The next day, when news of the rise comes out, netizens will definitely be condemning the state, CNPC and Sinopec. At this point, I register an ID and post a comment: “Rise, rise however you want, I don’t care. Best if it rises to 50 yuan per litre: it serves you right if you’re too poor to drive. Only those with money should be allowed to drive on the roads . . .”

This sounds like I’m inviting attacks but the aim is to anger netizens and divert the anger and attention on oil prices to me. I would then change my identity several times and start to condemn myself. This will attract more attention. After many people have seen it, they start to attack me directly. Slowly, the content of the whole page has also changed from oil price to what I’ve said. It is very effective.

What’s your area of work? Which websites do you comment on? Which netizens do you target?

There’s no limit on which websites I visit. I mainly deal with local websites, or work on Tencent. There are too many commentators on Sohu, Sina, etc. As far as I know, these websites have dedicated internal departments for commenting.

Can you tell which online comments are by online commentators?

Because I do this, I can tell at a glance that about 10 to 20 per cent out of the tens of thousands of comments posted on a forum are made by online commentators.

Will you debate with other people online? What sorts of conflicts do you have? How do you control and disperse emotion?

Most of the time we’re debating with ourselves. I usually never debate with netizens and I’ll never say I’ve been angered by a netizen or an event. You could say that usually when I’m working, I stay rational.

When the government says, “Don’t believe in rumours, don’t spread rumours,” it achieves the opposite effect. For example, when Sars and the melamine in milk case broke out, people tended to choose not to trust the government when faced with the choices of “Don’t trust rumours” and “Don’t trust the government”.

I think this country and government have got into a rather embarrassing situation. No matter what happens – for example, if a person commits a crime, or there’s a traffic accident – as long as it’s a bad event and it’s publicised online, there will be people who condemn the government. I think this is very strange.

This is inevitable, because the government encompasses all. When all honour is attributed to you, all mistakes are also attributed to you. Apart from targeted events, are individuals targeted? Would there be this kind of directive?

There should be. I think for the Dalai Lama, there must be guidance throughout the country. All people in China hate the Dalai Lama and Falun Gong somewhat. According to my understanding, the government has truly gone a bit over the top. Before I got involved in this circle, I didn’t know anything. So I believe that wherever public opinion has been controlled relatively well, there will always have been commentators involved.

How do your superiors inspect and assess your work?

The superiors will arrange dedicated auditors who do random checks according to the links we provide. Auditors usually don’t assess, because they always make work requirements very clear. We just have to do as they say and there won’t be any mistakes.

How is your compensation decided?

It’s calculated on a monthly basis, according to quantity and quality. It’s basically calculated at 50 yuan per 100 comments. When there’s an unexpected event, the compensation might be higher. If you work together to guide public opinion on a hot topic and several dozen people are posting, the compensation for those days counts for more. Basically, the compensation is very low. I work part-time. On average, the monthly pay is about 500-600 yuan. There are people who work full-time on this. It’s possible they could earn thousands of yuan a month.

Do you like your work?

I wouldn’t say I like it or hate it. It’s just a bit more to do each day. A bit more pocket money each month, that’s all.

What’s the biggest difficulty in the work?

Perhaps it’s that you have to guess the psychology of netizens. You have to learn a lot of writing skills. You have to know how to imitate another person’s writing style. You need to understand how to gain the trust of the public and influence their thoughts.

Why can’t you reveal your identity? Why do you think it’s sensitive?

Do you want me to lose my job? Whatever form or name we use to post on any forums or blogs is absolutely confidential. We can’t reveal our identity, and I definitely wouldn’t reveal that I’m a professional online commentator.

If we do, what would be the purpose of our existence? Exposure would affect not just me, it would create an even greater negative effect on our “superiors”.

What do you mean by “superiors”?

Our superior leaders – above that should be the propaganda department.

Is your identity known to your family? Your friends?

No. I haven’t revealed it to my family or friends. If people knew I was doing this, it might have a negative effect on my reputation.

You say: “If I reveal inside information, without exaggeration this could lead to fatality.” Do you think that the consequence would be so serious?

With my identity, I’m involved in the media and also the internet. If I really reveal my identity or let something slip, it could have an incalculable effect on me.

If you say you want to quit, will there be resistance? Are there any strings attached?

Not at all. This industry is already very transparent. For me, it’s just a part-time job. It’s like any other job. It’s not as dark as you think.

How many hours do you go online each day and on which sites? Do you rest at the weekend?

I go online for six to eight hours nearly every day. I’m mainly active on our local BBS and some large mainstream internet media and microblogs. I don’t work over weekends, but I’ll sign in to my email account and see if there’s any important instruction.

In daily life, will you still be thinking about your online work?

Now and then. For example, when I see a piece of news, I’ll think about which direction the superiors will request it to be guided in and how I would go about it. It’s a bit of an occupational hazard.

Do you watch CCTV News and read the People’s Daily?

I usually follow all the news, particularly the local news. But I generally don’t watch CCTV News, because it’s too much about harmony.

Do you go on Twitter? Who do you follow?

Yes. I follow a few interesting people, including Ai Weiwei. But I don’t speak on Twitter, just read and learn.

How big a role do you think this industry plays in guiding public opinion in China?

Truthfully speaking, I think the role is quite big. The majority of netizens in China are actually very stupid. Sometimes, if you don’t guide them, they really will believe in rumours.

Because their information is limited to begin with. So, with limited information, it’s very difficult for them to express a political view.

I think they can be incited very easily. I can control them very easily. Depending on how I want them to be, I use a little bit of thought and that’s enough. It’s very easy. So I think the effect should be quite significant.

Do you think the government has the right to guide public opinion?

Personally, I think absolutely not. But in China, the government absolutely must interfere and guide public opinion. The majority of Chinese netizens are incited too easily, don’t think for themselves and are deceived and incited too easily by false news.

Do you have to believe in the viewpoints you express? Are you concerned about politics and the future?

I don’t have to believe in them. Sometimes you know well that what you say is false or untrue. But you still have to say it, because it’s your job. I’m not too concerned about Chinese politics. There’s nothing to be concerned about in Chinese politics.

 

This article first appeared in the 22 October 2012 issue of the New Statesman, Ai Weiwei guest-edit

Photo: Getty
Show Hide image

Does the working class need to ask for its Labour Party back?

The more working class voters there were in a constituency in 2017, the more it tended to swing to the Tories.

When Theresa May called the general election nearly two months ago, all the evidence – opinion polls and local election results especially – pointed to the expectation that the Labour Party would be crushed, with many of its MPs losing their seats.

The assumption was that Labour under Jeremy Corbyn would be unable to win over Conservative voters, because he was too left-wing to appeal to those close to the political centre ground.

Some commentators, myself included, took this a little further, arguing that Corbyn was left-wing in a way that would alienate the very people he claimed to speak for, ie working class people, while appealing primarily to virtue-signalling middle class romantics like Corbyn himself, who have no more interest than he does in the business of parliament but love a good rally or social media spat.

The local elections that took place in May appeared to confirm the above expectation and analysis, with hundreds of Labour councillors losing their seats. However, opinion polls began to shift, and while different polling companies’ methodologies led to different estimates of support for the two main parties, all showed Labour on the rise – with YouGov predicting two days before the election that the Conservatives would win a mere 305 out of 650 seats, while Labour would win 266.

Despite a miserable campaign in support of a depressing manifesto, enlivened only by the promised revival of an anachronistic bloodsport beloved of the rural elite – indeed, a campaign so bad that political historian Glen O’Hara joked about having ‘watched and wondered whether Mrs May was a Corbynite sleeper agent’ – the Conservatives actually did slightly better than this prediction, winning their highest share of the vote since 1983 and coming to hold 317 seats to the Labour Party’s 262.

This left them only 55 seats ahead of their historic rival: a gap only very slightly wider than the 48-seat lead that they had after the 2010 general election, when David Cameron defeated the supposedly very unpopular Gordon Brown. The 2017 result would have been impossible without the activists who have stuck with the Labour Party regardless of their feelings about the leader, some of whom are now publicly expressing shame at the part they played in what is widely seen as Corbyn’s triumph.

Does the Labour Party’s unexpectedly narrow defeat refute the diagnosis of Corbynism as a middle class politics that alienates the party’s traditionally working class base, but doesn’t really care? Constituency-by-constituency analysis of the 2017 results by Paula Surridge, of the University of Bristol, suggests that it does not.

The Leave vote

We should perhaps begin with a pattern that was already apparent on election night. Parts of the country that voted strongly to quit the European Union appeared to show a swing away from Labour towards the Conservative Party, while areas that voted strongly for Remain appeared to show a swing in the opposite direction.* 

Surridge’s analysis confirms that this was indeed a trend: the higher the estimated Leave vote, the more the Labour vote share fell between 2010 and 2017, and the more the Conservative vote share rose during the same period. Blue dots represent actual constituencies; the red line represents the trend.

On the face of it, this is baffling. Both the Labour Party and the Conservative Party are officially committed to leaving the EU, and Jeremy Corbyn famously used a three-line whip to force his MPs to support the Tory Brexit bill in February.

The anti-Brexit parties were the Liberal Democrats, the Scottish National Party, and the Greens. There was therefore no sense in which a vote for Labour could have been a vote against leaving the EU. Why, then, should a constituency’s support or opposition to Brexit have made any difference?

This brings us to the paradox that the Labour MP John Mann has called the ‘Bolsover question’: why the second-largest Labour-to-Conservative swing in the country should have occurred in the constituency of Dennis Skinner.

Skinner is not only – as Mann observed – one of Jeremy Corbyn’s staunchest supporters in the Commons, but also  – although Mann did not draw attention to this fact  – one of the Labour Party’s staunchest advocates of Brexit. Why should a constituency that voted for Brexit by 29,730 votes to 12,242 have swung so heavily against a strongly pro-Brexit candidate for a pro-Brexit party?

Here’s a thought: maybe constituencies swung away from Corbyn’s Labour Party for the same sorts of reasons that they voted Leave, and swung towards it for the same sorts of reasons that they voted Remain? Or to put it another way: what if Corbynism appeals to the kinds of people to whom EU membership seems advantageous, and repels the kinds of people to whom it seems an encumbrance, regardless of the fact that Corbyn – as a disciple of Tony Benn  – is resolutely anti-EU?

Let’s take a look at some of the other things that Surridge found.

Educational level

Exit polling after last year’s EU referendum found that the more educated a person was, the more likely they were to have voted Remain. While some Remainers might like to dismiss this as ignorance on the part of Leavers, it can also be interpreted as an expression of anger at being left behind in Britain’s ever-more highly globalised economy.

So we should take note of Surridge’s finding that the higher the percentage of university degree holders in a constituency, the more it would tend to swing towards Labour from 2010 to 2017, and the lower the percentage of degree holders, the more it would tend to swing towards the Conservatives.

Ethnicity

While a bare majority of white voters opted for Leave last year, large majorities of black and Asian voters chose Remain. The reasons for this are complex – but it is notable that Surridge finds that the lower the percentage of white British voters in a constituency, the more it would tend to swing towards Labour, and the higher the percentage of white British voters, the more it would tend to swing towards the Conservatives.

While it is certainly good news for Labour that it is winning votes in more diverse communities, it should think carefully about why this is not happening in less ethnically diverse parts of the country – particularly as these are often economically struggling areas unattractive to immigrants.

Class

Now the biggest question of all. The Labour Party was set up to provide parliamentary representation for working class people, and the far left trumpeted Corbyn’s leadership as a triumph for "working class politics". But opinion polls showed something very different: under Corbyn, working class support for Labour rapidly fell to its lowest point ever.

Moreover, by-election results in the strongly working class constituencies of Stoke-on-Trent Central and Copeland showed swings from Labour to the Conservatives, as indeed they had during the Labour Party’s last flirtation with Bennism in 1983. Did the general election see working class voters change their minds and flock back to Corbyn’s "socialist" party?

My goodness. Surridge’s analysis shows that the more working class voters there are in a constituency, the more it tended to swing Conservative, and the fewer there are, the more it tended to swing Labour. To put some figures on that, she found that for every 10 per cent more working class voters in a constituency, there tended to have been a fall of about 3 per cent in the Labour vote and a rise of about 5 per cent in the Tory vote between 2010 and 2017.

Think about that for a moment. This is Corbyn bringing the party back to its "working class, socialist roots"?

Correlations, 2010-2017 and 2015-2017

I sense an objection: these figures show the swing from 2010 to 2017, and Corbyn’s only been in charge since 2015. Maybe it’s all Ed Miliband’s fault?

Apparently not. Surridge calculated the correlations between all the above variables and the change in the Conservative and Labour vote, both for the period of 2010-2017, and for the period of 2015-2017. And here they are:

While it is true that many correlations are weaker for the period 2015-2017 than for 2010-2017, the positive correlations remain positive and the negative correlations remain negative.

In other words, working class voters, voters not educated to college level, and voters in ethnically homogeneous areas love Corbyn’s Labour Party even less than they loved Miliband’s. Meanwhile middle class voters, those educated to college level or higher, and voters in ethnically diverse areas love it even more.

It should also be noted that the positive correlation between the percentage of working class voters and the change in the Conservative vote, and the negative correlation between the percentage of voters with degrees and the change in the Conservative vote, are both stronger for the period 2015-2017 than they are for 2010-2017, indicating a rapid growth of support for the Conservative Party among the very social groups that Labour traditionally represented.

This should worry Labour politicians with ambitions to be in government, because there is simply no way that a Labour leader can become prime minister without persuading Conservative voters in Tory seats to switch to Labour. Corbyn may have put together an unexpectedly large anti-Tory coalition of voters, but it’s largely concentrated in areas that already vote Labour – and traditional Labour voters are being driven faster than ever into the Tories’ arms.

The triumph of the "socialism fan"

In recent decades, Labour has become the party of anti-racism. It can be proud of the fact that its vote share has risen in ethnically diverse constituencies – although it seems to me that the racism many Labour supporters (and in some cases, activists and even politicians) have shown towards the Jewish community ought to be treated with rather more alarm than it apparently is.

But whatever the positives in this mixed achievement, it should be hard indeed for the party to find cause for celebration in the fact that the Conservatives are so rapidly becoming the party of the "left behind".

In the post-New Labour era – and even more so under Corbyn than under Miliband – Labour has become a party of highly educated middle class people, "socialism fans" especially. I said it before the election, and it remains the case today.

Indeed, the Labour leadership’s understanding of this point seems the most likely explanation for their manifesto pledge to end student fees (a policy that would benefit only higher-earning graduates, since people who do not go to university do not incur student fees, and people who do but end up in lower-paying jobs don’t have to repay their loans) while maintaining the Conservative "benefit cap", which negatively affects low earners, disabled people and the unemployed.

To what extent Labour’s new middle class voters will continue to back the party in the future seems unclear. After all, Corbyn can’t really do anything about their student fees, since he is not prime minister, and while he could do something about Brexit (since Labour, the anti-Brexit parties, and pro-EU Tories such as Ken Clarke now collectively hold a majority of seats in the Commons), he’s promised not to (good Bennite that he is).

Then again, he might publicly change his lifelong position on Europe just as he has publicly changed his lifelong positions on terrorism, nuclear weapons and Nato. He wouldn’t be the first leader to decide that Paris was worth a mass.

Fair play to him, though. In losing the election by only slightly more seats than Gordon Brown, he won the anticipated leadership contest in advance. So if the working class asks for its Labour Party back, he can confidently tell it to get lost.


* Canterbury is a notable exception here, having narrowly voted Leave in 2016 but swung to Labour in 2017. A very small city with two well-known universities, it hosts a very big student population during term time (when the general election took place), a large proportion of whom would typically have been expected to be resident elsewhere during the holidays when the EU referendum took place.

0800 7318496