Stanford Professor Jennifer Pan Discusses "Disguised Repression" in China
Aside from biweekly SCCEI China Briefs, we periodically feature relevant content from other SCCEI programs, like this lecture by Professor Jennifer Pan exploring her research on repression in China.
Below is a video and lightly edited transcript of Professor Jennifer Pan discussing her work on "disguised repression" in China summarized in the mid-October SCCEI China Brief: China Has Laws Banning Political Activism. Why Charge Critics with Non-Political Crimes?
Highlights of her talk are bolded in the transcript below.
Professor Jennifer Pan is the Sir Robert Ho Tung Professor of Chinese Studies, Professor of Communication and (by courtesy) Political Science, and a Senior Fellow at the Freeman Spogli Institute at Stanford University
Thanks so much for having me. Today, I'm going to talk to you about this joint work with Xu Xu, who's now a professor at Princeton, [but] who was a post doc [at Stanford] when we started this work, and Yiqing Xu, who's also here at Stanford.
In September of 2020, Ren Zhiqiang, a real estate mogul, was sentenced to 18 years in prison for corruption. But many people think that his real crime was criticizing the Chinese Communist Party and its top leaders. And his case is not unique. So many other critics of the Chinese regime have been arrested and sentenced for crimes unrelated to political activism. Xue Manzi, for example, and Ai Weiwei, among others. You see this around the world, where political dissidents and political opponents are sentenced for crimes unrelated to their political activity.
Shehbaz Sharif was the leader of Pakistan's main opposition party and was sentenced for corruption in 2020. Alexei Navalny, the former opposition leader in Russia, died while in prison for charges against him for embezzlement and bribery. Thaksin Shinawatra, after he was overthrown in a military coup in Thailand in 2006, was then charged with corruption. And the last example, Anwar Ibrahim of Malaysia, was leading the Malaysian opposition party until he was sentenced to 5 years in prison on sodomy charges in 2015. These are political opponents, political dissidents, who are imprisoned and charged with crimes unrelated to their political activity. And so, the motivating question for us is this: why do these authoritarian, and sometimes democratic governments, punish political opponents with these non-political crimes?
This is a puzzle because so many of these same governments have made many forms of political activity illegal. It'd be very straightforward to charge these same people with just political crimes. And if we step back a bit and look at the literature on repression and dissent, there's been a lot of research on the effects of repression, but these effects are totally varied. There is no consensus in this literature. Sometimes repression suppresses dissent, other times it catalyzes mobilization.
But regardless of the effects, there's this consensus that those who are targeted by repression don't like it. Right? They oppose repression. And the same of bystanders – everyone who witnesses repression happening is also in opposition to it. So, across this big literature on repression and dissent, generally, we think people are opposed to repression. And so, we're asking, is that always the case? What if, instead of framing it as repression, you describe the same activity through another lens, say, criminal justice? What happens to support or opposition for repression when that happens, when repression is portrayed differently?
I’m going to lay out our argument, then show you a survey experiment that we did, and then a case study for the external validity of this argument. Just to start, [I’ll review] some terminology to make sure we're all on the same page. When I talk about state repression, I'm talking about the actual or threatened use of physical or psychological sanctions taken by the state in response to some behavior – that's everything from, arrest, imprisonment, torture, and killings, to threat of those actions. Who are political dissidents? How am I using that term? We're using the term political dissident to mean people who collectively or individually challenge the political authority of those in power.
Dissident really differs by context. Many of you are very familiar with China and the Chinese political context. I'm using the word dissident here to talk about critics of the regime. So that might be different than what you typically think of as a dissident in China, but that's the definition broadly that we're using. And to be clear, a dissident in one context may be someone who is critical of the regime, but in another context, criticism is totally fine, and you would have to rise to some sort of armed insurgency to be considered a dissident.
But political dissidents are those who are challenging political authority in some way. They're more likely than the average person to take actions that challenge the political status quo, and so they're difficult for the regime to co-opt, and I would say repress. This is getting into the political science literature. We typically think that regimes can either co-opt, which is bribe and buy off, or repress, which is use coercion.
Dissidents are more difficult to co-opt because they're more in opposition to the regime. The cost of co-opting them will be quite high. And repression, as I mentioned, has really variable effects. When the Stanford administration tries to suppress protest, for example, let's say around Palestine, sometimes it works, other times it backfires. And that's what we know about repression in general – its effects are variable.
Because repression often backfires, and that's been especially true in the last 20 years or so, popular protests have become increasingly challenging for the survival of authoritarian regimes globally. If you think about the mid-1900 or 1940s-2000, most authoritarian regimes changed from coups rather than popular mobilization. That's kind of when I was in grad school, that's what we learned – that authoritarian regimes change when there are coups. But if you look at the data from 2000 to now, so in the last 20 years, twice as many authoritarian regimes have changed because of mass movements than coups. The relative role of coups versus mass mobilization has really changed in the last 20 years. From 2000 to 2017, about 60%, so more than half of all authoritarian regimes, face at least one large-scale, anti-government protest. And it's in this context, where there's popular mobilization happening under authoritarian rule, that when autocrats apply repression, that repression often backfires.
Why might that be? Well, protests are happening because of perceived injustice and illegitimacy of people in power, and then you have repression, or what we're calling blatant repression. Punishing opponents for their political activity, that means the regime is punishing opponents for challenging its illegitimacy, can then inflame those feelings of illegitimacy among opponents. What we think is happening instead of blatant repression is what we're calling disguised repression.
[Disguised repression] is the act of charging dissidents with non-political crimes completely unrelated to any political activity they're engaging in. If you think back to 2013, Xue Manzi was engaging in online criticism of the Chinese government and its policies online, but his charge was soliciting prostitutes and group licentiousness. He could have been charged with online criticism because laws were passed at that time, making certain types of online criticism illegal. But instead, the charges were what we call disguised repression, unrelated to his political activity.
This definition is defined by the type of charge, not by whether people believe it or not. It doesn't matter if people don't believe it or if people do. And we think it's advantageous for the regime to actually charge dissidents with non-fabricated crimes. So, I think in many contexts, there's the idea that you can fabricate a charge for a dissident. And I think that may happen, but if that's discovered, that would then reduce the effectiveness of this type of charge.
So why does it work? Why does it work to charge someone with disguised repression as opposed to blatant repression? We think the mechanism is through morality and moral authority. Morality is a set of fundamental standards about the distinction between right and wrong, good and bad behavior. And moral authority is some external constraint on behavior. And it's a resource that's been studied in mobilization and protests. And the dissidents serve, we argue, as role models that are expected to adhere to moral codes. When disguised repression happens, it damages those dissidents' moral standing. It paints dissidents as offenders who violate societal moral codes, lowering supports for the dissidents among those who follow them as well as bystanders. And then, justifying repression as punishment for immoral behavior – instead of repression as immoral acts on behalf by the state, instead, repression becomes just punishment for the immoral behavior of dissidents. As I mentioned, this definition is not about whether people believe or not.
But it doesn't matter whether people believe the charges or not. It creates cleavages among supporters. Some will believe the charges, some won't. And it decreases that coordination among these potential dissenters and reduces the willingness to dissent. We also argue that it induces self-censorship among other would-be dissidents.
So, when you're a dissident and you see someone else being charged with a non-political crime, that increases the potential cost of dissent. It means that if you're someone who might have engaged in behavior similar to those charges, you might then abstain from political dissent. Because now you realize the punishment is not just for your political activity, but for other forms of activity that you may have engaged in. So, we think it will induce self-censorship among dissidents.
Then the question is… are authoritarian governments or transitional governments/hybrid regimes using [disguised repression] all the time? And the answer is no. We don't think it's always being used because it has a cost. It requires time and effort to identify plausible or actual nonpolitical crimes. And the state has to follow the normal rules and procedures of its judicial system in order to be credible.
When you have a dissident that's being charged for tax evasion, it's not the case that most people who are being charged with tax evasion are dissidents. Most people charged with tax evasion are not dissidents. And so, when you charge a dissident with tax evasion, you're still following the normal rules and procedures, and going through the case as you would for anyone charged with this nonpolitical crime. Because of this time and effort that's required, we think it's only used against people with greater mobilization power. And instead, blatant repression, which is just punishing people for political crimes, sends a signal that certain types of political activity are not allowed.
We think there will be some combination of blatant repression and disguised repression in any authoritarian context. And there's some scope conditions. It requires that the judicial system exists and has certain level of capacity, so there are rules and procedures. There are prosecutors, who can move a case through the system, and that it has some independent credibility.
I think you'll see in my results that in the China context, China's judicial system does have this independent credibility. Because if the judicial system is seen completely as a vehicle for those who hold political power, then this charge working through the judicial system is not going to have any extra effect. So, these are the observable implications we think would accompany our argument: disguised repression is going to decrease support for the dissident, increase support for punishing the dissident, decrease other people's willingness to engage in dissent, and increase self-censorship among activists. And then dissidents with larger followings are the more likely targets.
That’s what we think based on the argument. And then I'm going to show you an experiment we did as well as a case study to illustrate some of where we think these observations are found. This is an online experiment, an online survey of about 1,000 people in a nationwide sample. That's a convenience sample. We screen people who are in mainland China who are over the age of 18. We measure their political attitudes pre-treatment. And then for a random subset of participants, we remind them that their responses are anonymous. That's our way of checking to see whether there might be social desirability or preference falsification.
Then we show them a description of a dissident and then randomly assign participants into four conditions: one in which they don't see any repression of the dissident, one with blatant repression, and two with disguised repression (soliciting prostitutes and tax evasion). Then we ask a series of attitudinal questions. We ask questions about whether they would dissent on behalf of the dissident. Then we measure self-censorship by asking a series of increasingly politically sensitive questions and seeing whether they answer those questions or select the option of “would prefer not to respond.”
These are some [more] pre-treatment measures: we ask the participants whether they have criticized Chinese government policies, rules, or regulations on social media or online forums. This is how we measure whether the respondent is a critic of some sort. Then we also have a pre-treatment measure of the respondent’s morality.
So, we're asking a series of World Value Survey questions around whether they think things like fraudulent collection of government benefits, stealing, prostitution, violence, etc., are never justified or always justified because we're trying to measure the moral standards of the participants. And why are we doing that? It's because based on our theory, those who have less stringent moral standards and who have engaged in criticism are the people most likely to self-censor when they observe disguised repression happening. Basically, if I'm someone who has less stringent moral standards, maybe I've done some bad things and yet I engage in criticism of the regime, when I see a dissident arrested for tax evasion or soliciting prostitutes, I might be like, “Oh, maybe I've done similar things – now I'm going to self-censor.” That's the theory behind it.That's why we have these pre-treatment measures.
Then we get into the criticism. What we show participants is this: we say a well-known opinion leader often criticizes current affairs and discusses viral public events online on behalf of the public. For example, calling for greater disclosure of officials’ property, fighting for equal rights to education for children of migrant workers in cities, or calling on government to strengthen law enforcement against child human trafficking. These remarks are widely disseminated on the Internet and have large social influence. This person is not a real person. These causes are a mixture of what other critics have expressed in China because we didn't want the participants to think of a particular person. We wanted the general idea of some person, but not a specific individual.
What are the treatment conditions? The pure control has no further information, just that description I showed you. The first treatment condition is blatant repression where the participant sees this: recently, this opinion leader, which you just read about, was taken away by the police. After a period of investigation and evidence collection, this person was charged with publishing and spreading a large amount of harmful information, which had an extremely bad influence on the public. The court sentenced the person to 3 years in prison.
That is the setup and this language… was taken from actual charges against real dissidents. This is based on the language that's typically used.
The second treatment of tax evasion… instead of publishing and submitting harmful information, now we have tax evasion and then the details of the verdict. In order to obtain more benefits, the person employed tricks to evade paying taxes, took advantage of some loopholes in the tax system. And then lastly, for soliciting prostitutes… this [language] was taken from Xue Manzi’s information about his arrest and sentencing.
We think the disguised repression is happening through the mechanism of morality. We randomly assign one third of respondents to measure their perceptions of the dissident’s morality. So, what do you think? So, they just saw the criticism, then they saw the treatment conditions, and then we asked, what do you think of this person's morality? Very low to very high. And we only asked a third of the respondents because we didn't know whether even asking this question would then influence the outcomes. So, we only ask a random subset about morality. And then finally, in our outcome measures, we're interested in attitudes toward the dissidents and attitudes toward repression. In terms of attitude toward the dissident, we're saying, based on the information currently available, what's your general attitude toward this person? Very supportive or very disgusted. Then toward repression, do you think it's reasonable to arrest this person? Totally unreasonable and totally reasonable. So that's two outcome measures of this attitude toward the dissident and repression. Then we also ask whether you're willing to dissent on behalf of the dissident. So, saying netizens are calling for solidarity with a detained key opinion leader, would you express solidarity on social media? Just yes or no.
At the end of the survey, we tell respondents they can answer a few more questions, and they always have the option of selecting I don't want to answer. These are questions that increase in their political sensitivity. I think the last question we ask is about term limits. But there are also other questions about policies. We define self-censorship as refusing, saying I don't want to answer this question or refuting the criticism and supporting the stance of the government. So that's the self-censorship outcome.
And this is what we find in terms of, attitudes toward the key opinion later… we're comparing disguised repression against blatant repression. What you can see is that when we show participants this description of tax evasion or soliciting prostitutes, it diminishes the support for the dissident compared to just the political crime. You'll also notice that even charging the dissident with a political crime decreases support. There's something else happening, which is not the focus of this paper, around just using the judicial system against dissidents that has a chilling effect. It also increases support for repression. We don't have the pure control here because they didn't see anything about the arrest. But comparing tax evasion and soliciting prostitutes against just the political crime, you see that there's greater support for the arrest, that more respondents think it was totally reasonable to arrest this person.
We also see a decreased willingness to dissent on behalf of the dissident. This is when you join others on social media calling for solidarity, and that decreases for tax evasion and soliciting prostitutes compared to the political crime. When we look at this, you know, among the one-third of the sample who answered questions about morality, you can see that those who saw the tax evasion or the soliciting prostitutes treatments evaluated the key opinion leader as less moral than those charged with a political crime, and all of these are less moral than someone who is not charged with any crime.
Then, turning to the self-censorship result, we find self-censorship among critics with less stringent moral standards. This is the full sample. And what we see is that there's no relationship between any of the treatments and self-censorship, willingness to answer those politically sensitive questions. But when we look at critics, you see that the prostitution treatment has a significant positive effect in increasing self-censorship. When we look at the less morally stringent critics, that effect is larger. The critics are people who said pre-treatment that they have criticized government policies on social media in China. And then the less morally stringent critics are people who were more tolerant of some of those World Value Survey questions on morality and were critical. So, that's an experiment, but it's only 1,000 people.
Is there external validity? Does what we're describing actually happen in the real world? That's where we turn to the case study of 2013 Weibo. Before people stopped using Weibo, Weibo was a place where people posted opinions on social and political issues. Thinking back to 2011, 2012, and 2013, people posted opinions on social and political issues, often challenging the narratives of the state.
Someone I've talked about before, Xue Manzi, had more than 10,000,000 followers on Weibo and often criticized Chinese government policies. But in the fall of 2013, there was a large crackdown on critics, especially online critics. We identified 28 critics who are arrested and charged during this period, some with political crimes and some for non-political crimes. There are 17 individuals who are charged with political crimes in this fall of 2013 period. The crimes include gathering crowds to disrupt public order, a legal assembly, picking quarrels, provoking trouble, etc. But then at the same time, there are 11 individuals who were arrested and charged with nonpolitical crimes. And these non-political crimes range from illegal business to profiteering to spreading rumors. That's like engaging in legal spam. Other charges include illegal dumping, extortion, blackmail, etc.
We had these 28 critics who are arrested and charged for political and non-political crimes. We searched for all Weibo posts from this 2013 period that mentioned their real names or their Weibo handles. And we collected about 13,000,000 Weibo posts that referenced these 28 people from 2011 to 2014. And this is the number of Weibo posts for those who are charged with political and non-political crimes…What you can see is that there was way more social media discussion of those 11 individuals who are subsequently charged with nonpolitical crimes. We see this as support for [our hypothesis]: the Chinese government is more likely to use disguised repression against influential critics instead of using blatant repression. But then when people are charged with disguised repression, dissent on behalf of the critic really decreases. You really don't see any dissent on behalf of the people who are charged with the non-political crimes.
[You can see from our data that] if someone is being charged with a non-political crime, no one is saying that they should be released, whereas for those 17 who are charged with political crimes, there are a number of posts that are calling for their release. Then we also look at how these critics are perceived in terms of support or antagonism before and after their charge and arrest. So, we have the Weibo post. We have the target, which is the name of the critic, and then we have a stance… we're [then] training a deep learning model, a BERT model, to identify stances using data.
…What you can see is that before the arrest, there's a pretty good level of support for these [dissident] individuals. There's also some opposition to these individuals. But then after the arrest, the against stance toward these people who are being charged with nonpolitical crimes really go up while the support decreases. And this looks very different than the political crimes. So, if you look at the period before the arrest, the kind of relative proportion of support and against is similar. There's more support, there's less opposition, there's some opposition. But then after the arrest, you see the switch for the non-political crimes, but the kind of opposite pattern for the political crimes. For political crimes, there's more support for these people after they're charged and arrested for political crimes. But that support goes away for those people who are charged with non-political crimes.
To sum up, when this repression is framed differently as immoral behavior, what we see is that there is support for repression and there's decreased support not for only for the dissident, but a negative evaluation of morality as well as support for dissent overall. And I think that has implications for our understanding of how governments use repression. In the last few years, we have we had a lot of research on how modern authoritarian regimes who are subject to greater media scrutiny can no longer can use blatant repression or no longer can hide repression in some way. But what we're finding is that instead of hiding repression, you can take the same actions, but you frame them differently in terms of the types of crimes that are being charged.
I think this has also implications for judicial systems under authoritarian rule. I also think there's been a lot of research talking about how when authoritarian regimes adopt rule of law, that it ties the hands of the autocrats. But what we're finding here is that those same autocrats can actually use a judicial system to serve their coercive purposes. I think there's kind of broader questions for future research around how both authoritarian and increasingly democratic governments are using law to achieve coercive or other types of political purposes.