Jennifer Pan, a 2019-20 CASBS fellow, is an associate professor of communication at Stanford University. Host John Markoff chats with her about her recent book “Welfare for Autocrats,” which explores how the Chinese government has reshaped a social assistance program into a tool of surveillance and repression.
Her recent book “Welfare for Autocrats”
Big thanks to CASBS staff member Teresita Heiser for opening the episode for us!
Narrator: From the Center for Advanced Study in the Behavioral Sciences at Stanford University, this is Human Centered. Today on Human Centered, Host John Markoff sits down remotely with Stanford Assistant Professor of Communications and Political Science Jennifer Pan. The two discuss her recent book, Welfare for Autocrats, which explores how the government in China reshaped its major social assistance program into a tool of surveillance and political repression.
John Markoff: So, um, you've just published a new book, Welfare for Autocrats, and I wanted to start there. And first of all, uh, you know, how can you have a book tour? How do you, how do you publicize a book in, in these new, in this new world?
Jennifer Pan: I have no idea. So this is my first book, so I have no reference for comparison. Um, so it actually came out in April, but then the publishers told me they had no idea when a physical copy would come out, um, so the Kindle version I guess came out in April, I was like, "I don't think my book's formatted at all for Kindle," so I waited, um, quite a while to begin talking about it, um, 'cause I also didn't feel that with everything happening in the US with Black Lives Matter, um, that it was necessarily the the right time to start talking about a book about China. Although there are some connections related to order or a government's obsession with order and what happens in that case.
John Markoff: Yeah, and also social welfare. I mean, it's the— I was very interested in the two systems. So I wanted to ask you about that, but I wanted to start by asking about your methodology. And could you— talk to us about how you did your research for this book?
Jennifer Pan: Yeah, so when I first started, this is what I wanted to do when I first started grad school in 2009. I'd worked for 5 years, mostly in China and mostly in healthcare. And I was interested in getting a deeper understanding of what the implications of government policies and government institutions were more. And so I decided to get a PhD after working professionally, which was not a traditional career path for what I'd been doing prior to my PhD. But one of the main things I was interested in was, and is, how do people live better lives? What can governments and civil society do to make people's lives better? And so this is the research I started with. At some point in grad school, I worked on propaganda, censorship, and that sort of, that research really took off. And so at one point I was ready to kind of give up this project, 'cause it seemed quite different from how my work had evolved. But I kept going 'cause this is what I was interested in. And I first started out by just doing fieldwork in China. So my main fieldwork was actually based in Wuhan, and Zhengzhou, which is the capital of Henan Province. So I spent months there getting to know people at the local level, grassroots levels. I then expanded my fieldwork, and it was actually after doing this kind of grassroots fieldwork that then I started gathering data more systematically to test some of the kind of intuitions and hypothesis that had come out of the on-the-ground interviews and conversations.
John Markoff: And so, um, well, two questions. One is, um, did you ultimately— you combined, um, uh, data that you collected with actual, uh, personal interviews you did with people? And, and yeah, that's— and were you able— what's it like to gather data sets in China? Do you gather them inside? Do you gather them outside?
Jennifer Pan: So This book consists of both. Most of my fieldwork happened in 2011, 2012, 2013. So that's when Xi Jinping was just coming to power. I think since 2012, 2013, China has changed a great deal. And the openness of, or the ability of outsiders to do research has changed dramatically. And so at the time I was doing fieldwork, it was still relatively straightforward to be able to collect data. Obviously, with local approvals and collaborations with local universities, people are open to having conversations. Now that's very, very different. And so, as you see in the book, a lot of the data also comes from these indirect sources, from social media, kind of extracting data from social media, other types of digital trace data. But I was lucky that I started my fieldwork when China was still relatively open.
John Markoff: And did you have experience, personal experience with surveillance when you were there? I mean, you're talking about a surveillance system, you know, to come back as an American citizen and spend time in China. How, what was it like?
Jennifer Pan: I think it's really hard. So I lived in China from 2004 to 2005, and then from 2006 to 2007, 2007-2009. Living in China, I just feel like this is such a vibrant, active place. There's so many people, why would anyone be paying attention to me? And that, that's the feeling you have as an academic. Every time before going to China, I'd be extremely paranoid. I don't bring my laptop, I don't bring my cell phone, I don't bring anything that has kind of my data on it. And I have a call-in protocol with my husband that's like, if I don't check in, then call the embassy and call these people, search for me. Nothing has ever happened while I was in China. I've definitely had people tell me that after a trip they were approached to ask about me, about what I was doing. What I was planning on doing. But I, and I've had collaborators invited to tea, you know, the euphemism for being asked to talk to security services, but I personally have not had that happen.
John Markoff: Okay. That's, it would probably be, as you said, it would probably be very different now, perhaps impossible to do the kind of research you did just 4 or 5 years ago.
Jennifer Pan: Perhaps. I don't know. I haven't been back to the kind of grassroots at the grassroots level to try to have these conversations. Yeah. Yeah.
John Markoff: Okay. You looked at, is it pronounced Dibao? How do you pronounce, sorry.
Jennifer Pan: So in Chinese it's Dibao or Dibao. So it's short, it's an acronym kind of for this minimum livelihood guarantee. But I think the anglicized pronunciation would just be Dibao.
John Markoff: Okay. And tell me, tell us a little bit about it and where does it fit in Chinese society?
Jennifer Pan: So Dibaoyu is China's main social assistance program, and what that means is that it's the government providing non-conditional cash transfers to households whose income falls below a certain level. So it's different than other cash transfer programs like Bolsa Família in Brazil, which are conditional. So in those programs, CASBS, the households have to do certain things in order to get the cash. In China, this program is, is supposed to be if anyone is below a certain level of income, then they get a certain amount of cash to bring them up to this minimum income level. In terms of actual expenditures, it's actually quite small when you compare it to something like pensions in China, but it's different than a pension, which is what I would call a social insurance program, because those programs are such that people and employers are contributing over time to these funds. This is something where the government is just providing cash to households.
John Markoff: And does it fit— you know, there's an international debate about basic income guarantees, and does it fit into that discussion in any significant way? Is it step toward that?
Jennifer Pan: So the idea is that everyone should have a certain level of income, and that everyone who falls below will be brought up to that level. And I think when the policy— so the policy was piloted in China in 1993 in Shanghai, and then rolled out nationally at the end of the 1990s. But over time, more and more requirements have been added to the program such that in some places it's not just looking at a household's income, it's looking at their assets, what, what they wear, whether they have phones or televisions in their houses. So over time it's become more of a traditional kind of means-tested program. So I would say it's maybe at the onset there was something that kind of has a origin that resembles where basic income debates have come from, but now it is really a means-tested kind of, um, traditional social welfare program.
John Markoff: And my, my understanding is that you, you came away— your takeaway that it is that it is now significantly used as a mechanism for social control. Is that—
Jennifer Pan: That's right. So one thing I should say is that the program— one thing that struck me with the program is that a lot of households whose income fell below that minimum income level did not receive the benefit. So this is what's called exclusion error. So for DBAL, 60 to 80% of households who should get the benefit don't get it. And this is based on research from China, so Chinese Academy of Social Science, National Bureau of Statistics. And so that led me to the question, well, why aren't these other households getting the benefit? And who's getting the benefit? And what I realized that there are kind of two logics that underlie who gets benefits. One is people who have no source of income. So not only is your income under this level, they're disabled, ill, elderly, in some way unable to participate in the labor market so that they could potentially gain income. And a lot of the households who were turned away had income below this level, but were deemed as able-bodied workers, that they can go out and find second, third jobs and make up that income. That's why they were turned away. But what I found is that in addition to kind of this population of people who could not work, getting debao, there were other people who could work but were also getting debao. And it happened that these people seem to belong to, um, a kind of— this is a Chinese government term— targeted categories or targeted populations. And these are people the regime thinks will threaten the regime somehow by protesting, by petitioning, by, um, belonging to certain religious organizations. And these people are prioritized for DIBAW.
John Markoff: So I see. So then what they're using is this is, this is a carrot. If you get out of line, you lose your carrot, you lose your support.
Jennifer Pan: Yeah, it's a very minimal carrot. So it creates this dependence. And one thing I want to kind of say is that I don't think policymakers at some point sat down, decided, and said, oh, hey, we have this DIBAW program. 'Let's use it for this population.' Instead, it's that when you look at policy documents from the public security— China's public security department— what they say is, 'Okay, we have this special population of people we want to control. Every possible resource that the government has that can apply to this population should be given to this population.' And so if they're low income, then we should put them on Dibao. So then the logic is really different because if you're a normal person coming to apply for D-Ball, you're like, you're told, oh, you look like you can work, why don't you just work? You don't need this benefit. But if you're in this population, they, they're, you know, aggressively giving you these benefits, and that changes the balance of, you know, where the resources go. That's what I'm referring to as seepage, kind of the Chinese government's really obsessed with what I call stability. And that seeps into how it works in many different areas.
John Markoff: And, and what is the, what is the role of the social media platforms in sort of watching, you know, who's in and who's out? Is, is it the principal mechanism?
Jennifer Pan: Um, I would say no. So the principal mechanism is at the neighborhood level. So China in urban areas are organized these neighborhoods, which are equivalent to rural villages. And in each neighborhood, there's a residential committee that's in charge of kind of organizing the administrative life of people in that neighborhood. And those who receive Diba, their names are actually publicized on bulletin boards in the neighborhood. So that's, I would say, the main way in which kind of this local community knows who is on Diba, who is not.
John Markoff: How To what degree is China urbanized and how much of this is, you know, if you, if you break it down urban, rural, how is it different?
Jennifer Pan: So this is something that is kind of changing quickly over time. A substantial proportion of the population in China is still rural. I think I have to look up what the number is. I want to say maybe like now 60% of the population population live in urban areas and 40% in rural areas according to official government definitions, but that's because in recent years there have been aggressive moves to try to urbanize previously rural areas. And part of this does have to do with welfare because the benefits that you get in urban areas are different than what you can get in rural areas and often much more generous in urban areas.
John Markoff: If you look at other socialist governments and societies, for example, Russia, when it was the USSR, the party was a mechanism for, for sort of delivering privilege. And so where does Dibao fit with respect to the party in China?
Jennifer Pan: So I would say that China probably moved— the welfare system in China dramatically changed after Maoist kind of institutions were dismantled. So if you look at, for example, healthcare from the 1970s to the 1980s, you see this huge shift from kind of public healthcare to completely privatized healthcare. And that really happened across many other areas. I think DBAO, when it was put in place in the late 1990s and early 2000s, was kind of the Communist Party recognizing that as it was pursuing these economic reforms and privatization, that it needed to provide some sort of social safety net, because there were a lot of protests around the dismantling of previous entitlements. But where the Communist Party is in the provision of Dibaoyi today, really, I think, comes back to that neighborhood level. So, according to, if you read regulations about Chinese neighborhoods, they're supposed to be, these residents' committees that administer these neighborhoods are supposed to be autonomous and grassroots. So, they're supposed to be somehow civil society. But they almost completely overlap with local party cells. So, the people who are running these residents' committees are often party secretaries and party cadres. So, at the grassroots level, the, So, in these urban areas that are the focus of my book, the Communist Party is really deeply penetrated.
John Markoff: You know, I asked you about the rural-urban dichotomy. I also wanted to ask you about aging in China and Dibo in terms of the elderly. I mean, I find China's demographic structure to be fascinating, and it's a harbinger for other things that are happening in the world, but particularly with respect to Dibo first.
Jennifer Pan: Yeah, so one thing I do wanna emphasize is that Dibao money goes to the poor and it is helping those who are poor. And the elderly, especially those who don't have family support, are recipients of Dibao. One thing that characterizes kind of China's welfare systems is that even throughout the Maoist era, family was always supposed to be the kind of primary safety net for a society. And so for the elderly, if those who are elderly have children and families to support them, that is seen as the primary mechanism still for supporting the elderly. And Dybbal goes to the elderly in cases where they have no children who can support them, where they're, they're without this sort of family welfare. Family social safety net. But I think that facing a kind of long-term issue facing the Chinese government is the fact that, you know, those over 60 years old are a significant proportion of the population. And with the one-child policy, um, there's a question of to what extent, uh, that aging population can be supported through this family social safety net alone. And I think especially now, given what's happening economically with COVID Yeah.
John Markoff: So, you know, in the United States, there was this, you know, the evolution of welfare in the United States toward workfare, if you will. The notion that you had to be employed. And, you know, you're also talking about Dibau, if you can get 3 jobs, you didn't need welfare. Parallels there?
Jennifer Pan: Or I, I think definitely if you— so, um, I think China in many ways, even though it has this communist history, is very liberal, has a very liberal, uh, kind of— I don't— that liberal doesn't mean the political liberal sense, but a liberal welfare system where individuals are supposed to take responsibility for their welfare. It's not the responsibility of the state. And so if you can work, you work. Yeah, yeah. And in China, I would say there's a lot of stigma against those who receive Dibao, because they're seen as getting handouts from the regime. And that's why you also, for those who are getting Dibao, in some ways it makes it easier for the state to control them because they are stigmatized and marginalized. But also it creates some backlash because when others see able-bodied people getting DBAO and they've been denied it, it leads to discontent.
John Markoff: I mean, is the framing of your book is that this is an essential— DBAO is an essential sort of structural element of the party control of the population? Is that the right way to view it or is it more complex?
Jennifer Pan: I think the way to view it is really about China's preoccupation with order has gone into every policy area, even a policy area that was really at root about social welfare and poverty alleviation, social security. Safety net. And I began, so I think ultimately it's thinking about kind of is, to what extent are these policies making people's lives better? And the intention of this policy absolutely is to alleviate poverty and to make lives better. But because of China's obsession with order, it's become diverted to these other purposes. And some people will have asked me, well, from the perspective of the Chinese regime, so from the perspective of the CCP, isn't it better that they're using this policy for control? And I'm not sure that that's the case. I mean, the counterfactual is not eliminating Dibao, but instead implementing Dibao as it was, as it is envisioned, which is to provide a social safety net. So the comparison is, what would China be like if the government gave Diba to everyone whose income fell below this level?
John Markoff: I also wanted to ask about your path to doing this kind of research. I saw this story that as a graduate student, you made this discovery about the Chinese government censoring social media posts. And I was wondering if you could tell us about how you stumbled across that aspect of the regime.
Jennifer Pan: Yeah. It began because I was working with another graduate student, now professor, Margaret Rob— Molly Roberts, and an advisor at Harvard, Gary King. We were working on developing a machine learning method algorithm to try to classify Chinese text. And we just needed lots of Chinese texts. So we were like, oh, let's collect social media data. A lot of text data there. And at some point during that work, which was just about political methodology, we decided to go back and check something. So we collected— in doing this research, we collected the URLs of social media posts and their text. At some point, we were like, we need to go back to the URLs to see if we're collecting the main text or ads or something like that. And when we went back, oftentimes we realized that the posts were not there anymore. So we're like, wait a minute, we have social media data, pre-censorship and post-censorship, and using that we can really try to understand what the Chinese government is censoring. And so that was really one of those by chance discoveries. I don't know how often that happens. My guess is that it doesn't actually happen like that very often, which led us down this path of studying government censorship and propaganda. And disinformation.
John Markoff: Moving towards social media, I saw you were quoted, I think it was in a Stanford Communications report on some of your work. You were quoted as saying there was no empirical evidence at all that social media leads to democratization, which I found very striking. And I was wondering if you could talk about that because, you know, mythically, And even today in America, there's this deep belief about democratization and social media.
Jennifer Pan: Right, right. So I still stand by that quote and I think it, because there are lots of layers to that. So there is empirical evidence that social media could help social mobilization. So in, so Maria Petrova, Ruben Nikolopov, and other colleagues have looked at social media in Russia and shown how, at least earlier on, use of VKontakte was probably, I mean, causally related to more protests and larger-scale protests. So I think that there's good evidence that kind of adoption of social media is correlated with social mobilization. But then you have to look at what's the relationship between social mobilization and democratization. I think that is much more murky. There's just so many other factors that relate to kind of the strength of grassroots organizations, civil society, to what level repression and kind of coercive capacity a regime has, what elites are oriented toward in that regime. So that, that's why I say kind of the linkage between social media and democratization That's not at all clear.
John Markoff: There was a period where there was reporting on these, you know, I guess labor-intensive surveillance systems that had been set up to watch social media in China. And I just wondered if you'd seen any indication that machine learning and AI is replacing human labor in terms of watching behavior on the internet. In China?
Jennifer Pan: [Speaker] So I would say that adoption of AI and machine learning is definitely happening and happening rapidly, but it's not clear to me whether it's replacing human labor or whether it's augmenting or acting simultaneous to human labor. 'Cause I think, sorry, sorry, on the censorship and control side, One of the things I've looked at is China's 50-cent army. So there are people who are purporting, supposedly posting on behalf of the Chinese government. And what we find is that people who are posting on behalf of the Chinese government are all government employees whose jobs have nothing to do with social media, but who are mobilized to make these posts on behalf of the government. Their jobs, the things that we observe them doing doing could easily be replaced by bots. But there's a question of whether it's in the interest of the regime to completely replace the activity these people are doing with bots. Because if their work is replaced by bots, it means that the regime is not controlling them or like mobilizing them and getting their compliance anymore. So yeah.
John Markoff: I also wanted to ask you about your time inside the Chinese CDC because it's become such a flashpoint in terms of the political discussion going on between the U.S. and China. Um, you know, uh, the Trump administration has accused the, the Chinese government of withholding data on COVID at the very beginning, um, uh, and CDC— the Chinese CDC has become part of that discussion.
Jennifer Pan: [Speaker] So I worked for the Chinese CDC right after SARS. This is in 2004, 2005. And I was working on HIV/AIDS, but working within kind of this part of the Chinese government made me realize that the central level government, so I worked for the central CDC, is constantly trying to get data from lower level CDCs, and they are met with resistance. Resistance. And I think what this reflects is the fundamental problem facing authoritarian governments. So you think of authoritarian governments as having a lot of control and able to get a lot of information. But one of the biggest difficulties for authoritarian rulers now and back in time is that their lower-level officials and lower-level agents have very strong incentives to hide information from the central level, especially because that information can reveal wrongdoing, corruption, mistakes. And so traditionally you have a secret police that is supposed to monitor not just the population but these lower-level bureaucrats. But without free media, these lower-level bureaucrats have strong incentives to try to bribe the secret police and collude. With any agent that's trying to collect information. And I think that's still true today. So I have one paper that looks at how even in the age of social media, when the central level government is trying to use social media to see what the public is really saying and, and saying about the local level governments, local level governments still have ways of hiding information from the central level. So when you get to a case such as COVID or whether something technical where you rely on testing, which happens at the grassroots level, this challenge of the center collecting information is definitely not straightforward. So I think when the US says the Chinese government is hiding information, that is actually obscuring a lot of the complexity within an authoritarian system where different people have information and there's not— there's no free flow of information. And when there's no free flow of information, people have incentives to hide information.
John Markoff: And, you know, writ large, the Chinese society seems to have done a better job, at least so far, in controlling COVID than our society. Do you have any impressions about why that is?
Jennifer Pan: So I think in some ways it's an— I mean, it's a society with authoritarian control, and when you have this authoritarian control, it becomes easier to control the population. Um, when you, you can impose quarantine strictly. Um, so I was talking to someone who has been doing research in rural areas post-COVID. And what he says is that immediately in hundreds of thousands of rural villages, everything was closed off. Everyone who returned from Wuhan clearly, or Hubei, clearly identified and quarantined. So in some ways, authoritarian government is well positioned to impose control, the type the type of control that's needed to stop a pandemic of this sort. But I think there's an open question of whether that means in the long term dealing with the consequences, it'll be similarly successful in dealing with the consequences of such a pandemic. I mean, I've been reading about kind of viral outbreaks and vaccination and talking about it with CASBS fellows, mostly in the tech group, talking about how when pandemics happen, in some ways it increases the power of the regime, of governments, because governments are then tasked with collecting information, enforcing certain types of behaviors. And that's the type of thing that was already in place in China because it has an authoritarian government.
John Markoff: On a different side, On that subject, the New York Times reported last week on a report by a group named Marco Polo, which basically argued that Chinese nationals have played and are continuing to play a key role in US technology companies. Particularly, they were looking with respect to artificial intelligence. I was wondering if you saw that article and had any general thoughts about the sort of state of US-China technology relations.
Jennifer Pan: I did not see that. Yeah, what was the— main argument there? Because I see—
John Markoff: argument, yeah, it was looking at the role that Chinese nationals who are here play and the research they've done and the impact of that research, I think, was the general—
Jennifer Pan: okay, okay. And is the idea that China's— US has benefited more from Chinese nationals than China?
John Markoff: Yes, that was it, that was the idea. Yeah, so the brain, the brain— yeah, what's the opposite of drain?
Jennifer Pan: The brain? I don't know. Yeah, I don't know. Yeah, yeah. So one thing, so I have actually a paper looking at Chinese students studying in the US and what happens to their attitudes when they're in the US. 'Cause one of the main motivations, I mean, I'm an immigrant, I was born in China, I immigrated to the US. My father came here to get his PhD in computer science and then ultimately stayed. Stayed and worked for US companies. But I think one thing that's been kind of personally, and I guess somewhat professionally relevant, is the increasing restrictions on research, on researchers, on scientists between the China-US. And I see those restrictions as not good for the long-term trajectory of either US-China relations or China's development.
John Markoff: Um, yeah, yeah, because perhaps the U.S. as well.
Jennifer Pan: I mean, uh, yeah, and what this would say with the— yeah, is the U.S. as well. Yeah.
John Markoff: Um, I, I just wanted to, to jump back to our discussion, uh, well, no, it's, it's sort of about Facebook. I just, I don't know if you've seen any evidence, but have you seen the reports that the Chinese government might be trying to play a role in the upcoming U.S. election in a similar fashion to the role the Russian government apparently paid in the last U.S. presidential election? Do you think that's a possibility?
Jennifer Pan: I think it would depend on similar in what ways. So I think if China sees that its interests are— that it can advance its interests by doing something with the U.S. election, then it might take such actions. But whether it will look like what Russia did in 2016, I think is an open question. I think there's also a big question of what China, what role China will play in the US election, separate from what China does to the US. So in other words, I think from the rhetoric that we see from at least Trump's party, political campaign. With coronavirus, China is definitely being positioned as the enemy, the adversary. Um, and so depending on how that plays out, that might also influence what actions China takes.
John Markoff: Neat. Thanks very much for spending the morning with us.
Jennifer Pan: Okay, cool. Well, thank you guys so much.
John Markoff: Yeah. Really appreciate it.
Narrator: That was Jennifer Pan in conversation with John Markoff about her new book, Welfare for Autocrats. We've got a link to the book in the show notes, and definitely check out her homepage there too. There are nearly two dozen papers exploring how governments behave online to influence political ideology and expression, so check those out. Human Center is a show from the Center for Advanced Study in the Behavioral Sciences at Stanford University. You can visit us online at casbs.stanford.edu or Twitter @casbsstanford. From everyone at CASBS, thanks for listening.