In Dictatorship and Information, Martin K. Dimitrov offers an approach to understanding possible solutions to the dictator’s dilemma (the balance between authoritarian governments’ use of information communication technology for economic development and the need to control the democratizing influences of this technology), which arises from the difficulty of calibrating repression and concessions due to distorted information about elite and popular discontent. Dr. Dimitrov argues that communist regimes are adept at developing sophisticated systems that mobilize the party, state security, and internal journalism to assess levels of dissent. Drawing from evidence across multiple communist regimes and numerous interviews, Professor Dimitrov contributes to our understanding of how autocrats learn – or don’t learn – about the societies they rule, and how they maintain or lose their hold on power.
In an interview conducted on March 7, 2023, Martin Dimitrov discusses information-gathering of communist regimes and how it extends authoritarian governments’ tenure with Kellee Tsai.
KELLEE TSAI: Hello. I’m Kellee Tsai, the dean of humanities and social science at Hong Kong University of Science and Technology. On behalf of the National Committee on U.S.-China Relations, I am excited to be speaking with Martin Dimitrov, professor of political science and chair of the department of political science at Tulane University, about his new book, Dictatorship and Information: Authoritarian Regime Resilience in Communist Europe in China, published by Oxford University Press.
Martin, I have the highest respect for your research on the study of comparative communism and socialist regimes, and I’m still referring students to your edited volume entitled Why Communism Did Not Collapse: Understanding Authoritarian Resilience in Asia and Europe, which was published by Cambridge University Press in 2013. It’s hard to believe that it’s already been a decade. You’re one of the few China scholars in political science who consistently brings a comparative perspective to our understanding of China. And I have learned so much from your work over the years.
So when Margot Landman at the National Committee invited me to interview you about your most recent book, I immediately accepted. But I had no idea that I was signing up for a magisterial 450-page plunge into the richly detailed mechanisms of how communist dictatorships have attempted to get accurate information about what citizens are really thinking. But it was a genuine pleasure to take that plunge, and I hope our listeners will do the same and read your book.
One of the central concepts in your book is that of the dictator’s dilemma. What is the dictator’s dilemma?
MARTIN DIMITROV: Kelly, first of all, thank you so much for agreeing to join this event and to talk with me about my book and also this volume that you mentioned, Why Communism Did Not Collapse, which indeed was published ten years ago. One of the great pleasures was to have a contribution from you in that volume. So we do have a chapter from Kelly Tsai in that volume.
Now back to dictatorship and information of the dictator’s dilemma. The dictator’s dilemma is the inability of dictators to have accurate assessments of the level of popular discontent that they face from society. And the argument is that because dictators are unable to accurately assess levels of popular dissatisfaction, they are fundamentally insecure and they’re likely to experience unanticipated coups and revolutions.
TSAI: That is a dilemma. So what are some of the main strategies that autocracies pursue to mitigate the dilemma?
DIMITROV: Right. In the classic literature, the argument is that the dilemma cannot be solved. So what dictators do is they engage in more and more repression, but more and more repression only makes them more insecure. So in the standard of understanding of the dictator’s dilemma, because it is unsolvable, it only increases the likelihood of coups and revolutions.
But what we observe when we look into actual authoritarian regimes is that some of them last a very long time. And in fact, there is one subtype of authoritarian regimes, the single-party communist regime, which outlasts every other type of dictatorship. So my book focuses on those regimes, in particular on single-party communist regimes. And I argue that they develop a whole range of mechanisms for alleviating and potentially solving this dictator’s dilemma.
And there, I distinguish between two main mechanisms. There are some institutions that allow citizens to voluntarily transmit information to the regime, primarily in the form of citizen complaints. And then there are other mechanisms that enable autocracies to involuntarily extract information, and those are primarily in the area of surveillance, but the range of mechanisms is extremely broad. So we have a rich panoply of institutions that authoritarian regimes create over time to mitigate this problem of the lack of knowledge about levels of popular dissatisfaction.
TSAI: Well, speaking of information, before we go into more detail about your argument, I’m curious about your sources of information. This is an especially difficult topic to research because information gathering in autocracies is a “hidden activity,” as you write. And it’s just remarkable that you conducted nearly 100 interviews in China, Cuba, Russia, Germany, Bulgaria, and the U.S. In addition to all of those field interviews, you also use the methodology of, quote, “archival ethnography.” Can you explain what archival iconography means in practice? What types of sources did you consult?
DIMITROV: Yes. Thank you, Kellee. I had a lot of fun writing this book and the fun involved trying to understand precisely what you mentioned, these hidden activities and authoritarian regimes. What do autocracies do to solve this problem? So, I started out by examining archives that have been reopened in Eastern Europe after the collapse of communism. Both the Communist Party and the state security archives are now open to researchers, and I also looked at the archives of the Telegraph agency. Those are the three main collectors of information in any communist regime.
And then I started engaging with Soviet archival materials which are available both in Russia, but actually more so outside Russia in various repositories around the world.
And then the big question was, where can I find equivalent Chinese sources? So, after a lot of searching in libraries in the U.S., in China and also in Hong Kong, I collected what I thought was an equivalent set of materials on China.
Now, archival ethnography is this method of reading materials not just as machine readable text, but actually asking, who were these materials produced for, with what purpose, and what might they be saying, in addition to the actual content of the document? They’re often marginalia, so the instructions from leaders, their reactions are to this, to this document. So for me, the question was who’s this document prepared for? How does it circulate?
And answering these questions, I argue in the book allows us to understand this problem of information flow in autocracies much more fully than we would if we were to subject these materials to computer analysis, for example, and treat them as machine readable text.
TSAI: Thanks. The bulk of the book is structured around pair chapters that compare Bulgaria and China during parallel phases of their communist governance, until Bulgaria transition to democracy in 1989 to 1991. What motivated you to compare Bulgaria with China?
DIMITROV: Well, I could give you two sets of answers, and one is the honest answer that I was born in Bulgaria, I grew up there, and I was 14 when ‘89 came along, and I didn’t understand what happened. And part of my reason for studying in China is that China gave us a communist regime that also experienced ’89, but the communist system persisted after ‘89, whereas in Bulgaria it didn’t. So, I have been thinking about Bulgaria my entire life.
But the scholarly answer, of course, is different from the personal one. And I argue in the book that Bulgaria and China are an unusually good pair for comparison, because I’m interested in two states that are unitary that did not have foreign troops on their soil. A lot of East European countries had Soviet troops, Bulgaria did not. I also wanted to regimes that are ethnically heterogeneous and have an ethnic minority population of about 10%. So, Bulgaria and China were a very good set of countries to compare. And I also wanted two countries where repression declines over time, and this allows for the voluntary provision of information.
So in in the first chapter of the book, I have a scholarly reason for comparing Bulgaria and China, and I make an argument that they are an especially good set. And if one were to compare China and the Soviet Union, which a lot of scholars do, they’re not as comparable as Bulgaria and China because there are a lot of other factors that vary between the two countries.
So, yes, so there’s a there’s a personal reason and there is a scholarly reason. And then in the case of Bulgaria, I also had archival materials available to me and I could collect them and study them and then find out ways for collecting similar materials from China.
TSAI: Well, both the personal and scholarly reasons make a lot of sense to me, for sure. I’d like to follow up on that distinction between involuntary and voluntary sources of information in communist regimes. Could you provide some examples of involuntary versus voluntary types of information? And is one type of information more useful for regime resilience than the other? And if so, why?
DIMITROV: Yes, this is a great question that requires a lengthy answer, but I will try to provide a brief one in terms of voluntary provision of information.
The one channel that I focus on the most is citizen complaints, and what I argue about citizen complaints is that when citizens complain, they present the government with requests for services and benefits like housing or jobs, or the payment of their wages or pensions. And when they make these requests, they don’t lie. So, citizen complaints are free from preference falsification, which involve citizens lying to the government about levels of support, so they are a very reliable mechanism for assessing levels of dissatisfaction. So this is one channel for voluntary transmission.
In terms of the involuntary, we have surveillance of individuals, and this surveillance can be conducted by state security, by the police, by the Communist Party, by other actors. There are many actors who can engage in surveillance. And the problem with surveillance is that there is a non-negligible probability that citizens would be aware that they are placed under surveillance, so they may misrepresent their support for the regime. So the problem with the involuntarily extracted information through various surveillance mechanisms is that it presents a greater hurdle for the analysts who are interpreting it. And of course, it is quite useful. And communist regimes engage in massive involuntary collection of information. But interpreting that information is less straightforward than the one that is voluntarily provided to the regime.
TSAI: Your book includes a table that lists 14 different mechanisms for assessing discontent, including techniques that are a bit unusual, perhaps, like monitoring offline rumors, monitoring offline jokes, and even monitoring dreams. Can you tell us a little bit more about those unusual mechanisms?
DIMITROV: Yes, thank you, Kelly. I mean, they’re certainly unusual, and they’re fascinating. I mean, I thought, whoa, so we have evidence in both China and Bulgaria and East Germany that there’s this monitoring of rumors. In East Germany, there was a systematic collection of jokes, anti-regime jokes. A similar effort took place in the Soviet Union and in Bulgaria.
And in terms of dream monitoring, I have there’s only one single report that actually comes from Bulgaria of state security, in the case of Bulgaria, collecting information on what people dreamt about. So these are unusual mechanisms, and the problem for the regime insiders is that this information was not particularly granular. They didn’t know what to do with it. And as a result, these mechanisms never emerged as the main mechanisms for assessing levels of discontent. Because, yes, I mean, jokes spread very fast and they’re often funny, but it’s not clear how subversive an individual who tells a joke actually is. Sometimes people just like to have fun and they tell jokes against the regime.
TSAI: Well, it seems that dictatorships often don’t have a very good sense of humor.
DIMITROV: Certainly true. (laughs)
TSAI: Jokes involving the dictator. (laughs)
DIMITROV: Well, although in the case of Bulgaria, the dictator, Todor Zhivkov, a man who was in charge for 35 years, asked his closest associates to tell him about jokes that people were spreading about him. And these associates say that he laughed. I don’t know how reliable that is, I mean, that’s memoirs. I tend to be very dismissive of memoirs, I don’t have archival evidence that he laughed about when he was told about jokes about himself, it’s just memoirs. But he did ask by the 1960s and then 70s and 80s that individuals who spread anti-regime jokes should not be prosecuted. So, I mean, at some point, you know, in that one instance, at least, the dictator decided that you know, you’re not going to put people in jail because they tell jokes about him. Now, other dictators are perhaps more sensitive about humor directed against them.
TSAI: Absolutely. Absolutely. The emperor doesn’t want to hear that he’s naked or that it’s funny that he’s naked.
TSAI: Your book shows that in the case of Bulgaria, it’s possible to solve the dictator’s dilemma. But that doesn’t guarantee that the dictator will be able to stay in power. Why is that?
DIMITROV: That’s paradoxical when we first hear it. Because the whole argument in the literature has been framed around the impossibility of solving the dilemma. And in the book, I show that, in fact, the dilemma can be solved. Autocracies are perfectly capable over time when they make the proper investments in the institutions to collect information about popular discontent.
But what I argue is that this information needs to be put to good use, and the good use from the point of view of the dictator is to engage in strategic redistribution and then targeted repression. And this strategic redistribution and targeted repression is only possible under certain conditions. One condition in my book is high levels of economic growth. Those allow for the redistribution. Another condition is a low, I mean, moderate, levels of fear. So when fear is very low, then repression, they need to repress too much. While levels of economic growth are too low, they don’t have enough funds for redistribution.
So in the case of Bulgaria, especially in the 1980s, what we had was excellent information. The government knew extraordinary details about how popular and how unpopular it was. But we had a lowering of economic growth. There was an economic crisis in the making, which meant that this strategic redistribution was compromised. And then levels of fear had declined to very low levels, which meant that the regime could not engage in targeted repression—it had to engage in mass repression. It just didn’t have enough repressive capacity.
So ‘89 was foreseen in Eastern Europe. The regimes knew that their popularity was eroding very fast, and they were unable to act on the information that they had. So for me, ’89 is not a story of lack of information. It’s a story about incapacity to act on the information that was available.
TSAI: That’s a really interesting insight. I’d like to switch gears to China. How have China’s information collection strategies changed over time? And more specifically, what are the implications of the regime’s use of more technology-intensive forms of surveillance these days? For example, are public security, police, and informants being automated out of their jobs?
DIMITROV: Yes. Thank you, Kelly. This, of course, is a question that has extraordinary importance for China.
What I show in my book is that in contrast to this stylized fact that circulates about East Germany being the most densely penetrated country in terms of the rate of informants, I actually demonstrate that China had more secret police informants than East Germany. And this continues despite the proliferation of very sophisticated surveillance machinery.
And, you know, one question that that I’ve gotten in the past is, well, why does China need all these humans that are secret police informants, given the extraordinarily high density of cameras? Given this vast apparatus that is reading people’s mail and their WeChat messages and so on and so forth? And the argument that I’ve made is that the information that is collected through the cameras and through the pervasive monitoring of social media is not sufficiently granular. It’s oftentimes unclear whether individuals have subversive intents, and what the government needs is it needs other humans to help interpret the information that is automatically collected through the cameras and through the monitoring of social media.
So humans continue to be extremely important in the information collection infrastructure in China today, despite the advent of the cameras and the algorithms that allow the regime to read social media.
TSAI: One of the key themes in your books concerns the challenges of gathering information from ethno-religious minorities in concentrated regions. I mean, that’s really a perennial challenge for dictatorships with ethnic minorities, so communist regimes have tended to opt for repression. Are there any strategies that would be more effective than the more coercive options?
DIMITROV: Yes. Thank you, Kelly. This, yeah, this is a very important question.
So what I do in the book is I compare the ethno-religious minorities in Bulgaria and the ethno-religious minorities in China. And I find that—and then there are side glances at the Soviet Union, which of course, was barely Russian—it was 49% ethnic minority.
So, the problem of penetrating the ethnic minorities is a problem that communist regimes prove to be unable to resolve over time, and they indeed opt for repression and for ethnic assimilation. And the reason why they do that is that they’re unable to recruit enough ethnic informants, and they also have a very low voluntary provision of information through the complaints system.
So the strategies that would allow the communist regimes to avoid the massive use of repression and resorting to assimilation techniques would be to incentivize citizens to complain. So more voluntary provision and to find ways to recruit more ethnic informants in these ethnic minority areas so then they can protect the two systems of the involuntary and the voluntary provision of information and avoid resorting to ethnic assimilation.
But unfortunately, the experience of both Bulgaria and China indicates that this has proved impossible over time.
TSAI: I really appreciated the insight in your book about how citizen complaints, the voluntary provision of information, is actually a good and healthy thing, that it actually reflects trust in the regime’s potential capacity to address those issues. As someone who’s serving in an administrative capacity, I you know, every day I open up my inbox and I worry that there might be one of those emails, one of those citizen complaints sitting there. But maybe I should feel assured when we receive these sorts of letters and visits, as it were.
One thing that surprised me about your book, and I love to be surprised because being in this field for a long time, I kind of feel like, oh, you have the general idea of what’s going on in China. But I did not realize that China actually has the lowest rate of citizen complaints compared to its communist counterparts. That’s quite remarkable. Is that something that you expected going into the research?
DIMITROV: No. Although I want to comment first on this issue of how wonderful it is to have complaints, because they show the trust of the populace. And Kelly, you’re a dean. I’m a department chair. But in a way, whenever you assume one of those positions, then instead of just studying complaints, you actually have to handle them. You receive complaints and then you have to resolve them. So, I have found that even though I’m fascinated by complaints from a scholarly point of view, handling complaints is quite different from studying complaints. But, you know, that being as it is. (laughs) Yes. So these days I like you. I encounter complaints—on a weekly basis, not daily.
But now the issue of China and the rate of complaints. I did not have a prior expectation—but I suppose I should take that back because, of course, the literature has been arguing that China is a very contentious place—but in the literature, there is a—protest and complaints are considered to be the same by certain authors. So, the fact that China has a lot of protests, a lot of scholars have argued that, of course it has to have a lot of complaints.
I find those to be substitutes. I mean, they’re not complimentary goods. So when we have more complaints, we have fewer protests. And when the rate of complaining goes down, we have more protests.
What I did expect when I went into the archives in Eastern Europe is that as the regimes were getting close towards collapsing and levels of dissatisfaction increase, the number of complaints would go up, but in fact it went down. And then what I found in the archives was that communist regime insiders interpreted the decline of complaints as an erosion of trust. This was in in Eastern Europe, and then I found a very similar interpretation of the relationship between a high volume of complaints and trusts. A high volume of complaints—and we have this from none other than Xi Zhongxun, Xi Jinping’s father—this is a quote from him, that if we have more complaints, that means that more people trust us.
So, an increase in the number of complaints indicates an increase in trust. And then when levels of trust decline, citizens go to the streets. So this emerged from the archives. And I think this relationship between complaints and protest came as a surprise to me because the China literature oftentimes thinks that complaints and protests are the same, but in fact, they move in opposite directions.
TSAI: I completely agree that they are they are quite different from the perspective of the ruling party. You’d much rather have people express themselves in words than heading out to the streets. Yes.
DIMITROV: And in fact, we have this most recent government restructuring, which was just announced yesterday, where the National Bureau of Letters and Calls has just been upgraded to being directly subordinate to the State Council, which indicates that even today, in 2023, the Chinese government would prefer for citizens to express themselves through the complaints system rather than to go out into the street and engage in a protest.
TSAI: I’m going to ask a question which may not be fair. Although your book is clearly and expertly addressed towards readers interested in understanding the information collection methods used by dictatorships, it could also be read as a handbook for dictators, i.e. how to collect information more effectively. I think there are policy implications, if not lessons, for contemporary or aspiring dictators.
At the same time, I think there are also potential lessons for aggrieved citizens. So, I’d like to focus on that piece. Could you share with us what some of the takeaways are for dictators—or really for those living in single-party communist regimes on the other side?
DIMITROV: Well, thank you, Kelly, for asking me about citizens rather than lessons for dictators, because others have also asked me, well, who’s this book for? Is this a handbook for dictators?
The book emerged out of my desire to understand how these regimes function. And in terms of citizens, one of the arguments that I make is that communist regimes care about what citizens think, and communist regimes want to incentivize citizens to participate in the complaints system.
So if there’s any lesson for citizens, is that they should engage in more contacting of the government and they should let the government know what they’re unhappy about. And in fact, communist regimes are more responsive to citizen complaints that than we often think. And long-lasting communist regimes realize that they need to respond to citizen input in order to stay in power even longer.
So if we are focusing on a positive lesson, I think that’s probably a positive lesson for citizens in autocracies that they need to be more engaged with the regime.
TSAI: Well, one potential response to that is there could be serious repercussions to issuing complaints, though.
DIMITROV: Correct. There are certain areas which are off limits. So, a lot of these complaints are about socioeconomic matters. And, in fact, if one were to complain about strictly political matters, then indeed, there are repercussions. There is also the question of, where does one complain? There has been retaliation.
But on that front as well, the longer a communist regime persists, the more attuned it becomes to the importance of protecting complainants from retaliation. Because it knows that if complainants are afraid of retaliation, they will lose that trust in the government and they will be more likely to turn to the streets and to engage in protest. And that, in fact, is one thing that communist regimes do not want.
They, of course, learn from protests. And on that as well, there’s an extensive literature on the information on value of protests. And there I argue that, of course, protests transmit information. But I personally have not found evidence that communist regimes encourage protests, even though they learn from them. In fact, they do quite the opposite. They actively discourage protests, and they draw the requisite lessons.
So if the communist regime wants to survive, it needs to incentivize its citizens to complain, and it needs to be sufficiently responsive to those complaints, and it needs to protect citizens from retaliation. So I suppose here there are lessons both for the dictators and for the citizens.
TSAI: Well, that’s a good segue to my final question to you, which is, the first sentence of chapter one in your book starts with, quote, “Eruptions of discontent are often unexpected.” Just a few months ago, we were all shocked by the so-called blank A4 paper revolution in China, which was followed by the complete lifting of COVID restrictions. Could you tell us how the insights from your book could be applied to explaining what happened?
DIMITROV: Thank you, Kelly. I like archives, and on this particular issue, we don’t have the archival evidence yet. So I think one answer would be maybe in 5 years I can tell you, or 10 or 15 what exactly happened with these protests.
But in terms of the argument that advance in my book is that in order to be resilient and to remain in power, the government needs to be responsive to popular input. So we can think of the A4 revolution as one instance in which citizens made clear what they wanted, and the government responded. Now, of course, what we can think of is that the government by then should have known full well what citizens wanted and it should have responded a lot sooner. So, one could certainly argue that, you know, responsiveness that comes at such a late stage is not what citizens wanted, and when we can get into all of the issues about zero-COVID and to what extent this policy reflected the lack of information versus, you know, other problems in the top leadership.
But I think in one way to think of the A4 revolution is that people spoke, and the government responded.
TSAI: Great. Well, Martin, thank you so much for sharing your insights with us. I hope this podcast has inspired listeners to run out and get a copy of your book. It’s just such an impressive undertaking that you have accomplished, and I think your contribution is going to be, you know, taking its legacy on our bookshelves and our reading list for years to come.
So I encourage everyone to check out Martin’s latest book, and thank you for joining us.
This transcript has been lightly edited for clarity. Please refer to the video interview to ensure accuracy.