top of page
  • Admin

Martin K. Dimitrov on China's Regime Resilience in Comparative Perspective

Tulane University Professor Martin K. Dimitrov talks to CACR about his book Dictatorship and Information: Authoritarian Regime Resilience in Communist Europe and China. This interview has been edited for length and clarity. (October 3, 2023)

Martin K. Dimitrov is Professor of Political Science and Chair of the Department of Political Science at Tulane University. He obtained his BA in 1998 from Franklin and Marshall College and his PhD in Political Science from Stanford University in 2004. His books include Piracy and the State: The Politics of Intellectual Property Rights in China (Cambridge University Press, 2009); Why Communism Did Not Collapse: Understanding Authoritarian Regime Resilience in Asia and Europe (Cambridge University Press, 2013); The Political Logic of Socialist Consumption (Ciela Publishers, 2018); and Dictatorship and Information: Authoritarian Regime Resilience in Communist Europe and China (Oxford University Press, 2023). Currently, he is completing a book entitled The Adaptability of the Chinese Communist Party. He was a fellow of the Public Intellectuals Program of the National Committee on United States–China Relations.


For the benefit of our readers who haven’t had a chance to read your book yet, do you want to tell us a little bit about what you looked at in this study and what you found?

The book engages with a prominent theory in political science which argues that dictators are unable to collect information on the level of popular opposition that they face, and as a result, they are unable to calibrate the two major tools of governance that they rely on. One is repression and the other one is concessions. And then this theory argues that because of this inability of dictators to assess the level of popular discontent, dictatorships are brittle regimes; dictatorships overrepress and underdistribute, and as a result, there’s an empirical expectation that dictators will experience regime collapse quite often.

So for me, the starting point was that we have a subtype of authoritarian regimes, namely single-party communist regimes, which outlast any other type of dictatorship, and this suggests that they have perhaps found ways to mitigate or maybe even solve this dictator’s dilemma. So what the book does is, it documents the extraordinary effort that communist regimes put into developing mechanisms for the collection of information, and then discusses the conditions under which this information is used to inform governance decisions. The book also makes an argument about how information helps extend the lifespan of dictatorships, but in the end, I argue that even with perfect information, a dictatorship cannot last forever. Dictatorships that collapse, the ones in Eastern Europe, did not collapse because they didn’t have information, I argue; they collapsed because they couldn’t act on the information that they had. So not having information has negative consequences for dictatorships, but sometimes having a lot of information doesn’t help them either. There’s this optimal level of information; the conditions have to be just right. Enough information has to exist, and dictators have to be able to use it to govern, and then under those conditions, information extends the lifespan of dictatorships.

That’s the argument in brief, and it’s based on a long process of the collection of primary materials, and the book says certain things about how these materials can be used and what methods we can employ to make sense of this information that the book is based on.

There’s something else I wanted to ask you about for the benefit of our readers, and that’s archival ethnography, which is the approach that you talked about taking in your book. Can you tell us a little bit about why you took that approach, why it’s important, and how it allowed you to learn some things and find some answers that hadn’t been uncovered in the past, including in cases where scholars had looked at some of these questions before?

Yes, this is an important question about the methods that I have used in the book. I do use other methods beyond archival ethnography– I conducted almost 100 interviews with various individuals in 7 or 8 countries, and I read a lot of memoirs– but the main source of data was documents that were prepared for regime insiders. When these documents were generated, the individuals who produced them did not have any expectation that they would circulate beyond the hands of their intended recipients. So what these documents allow us to see is how the regime talks to itself, how insiders talk to their bureaucratic superiors, and what information they transmit about the sources of popular discontent and the various measures that can be taken to counteract this discontent.

This process that I engaged in, which involved collecting the materials and then processing them and then thinking about them, is very similar to the process of ethnographic immersion that scholars undertake when they conduct standard ethnography. Standard ethnography involves going to a place and immersing yourself into the communities that inhabit that place with the hope that through this process of immersion, which takes place over time, you would gain some type of understanding about the political processes that unfold in this community. Writing this book took me a very long time. It involved work in several dozen archives in multiple countries over a 15-year period of time, so there was a lot of time that I spent going to these archives and the countries that they’re located in, and figuring out what the archives held, and collecting materials, and then initially reading them in the archive but then subsequently processing them at home and then writing about them.

For me, at the beginning of the process of researching this book, information collection was a black box. I knew very little about the mechanisms that dictators used to collect information, and at the end of the process of writing this book, I felt that the box had begun to open somewhat. And this is a metaphor, but to use this notion, I felt that I was gaining some understanding of the logic that was driving this process of information collection, transfer, and use.

The depth and breadth of the archival research was extraordinary. I felt like this was really a model study. On another point, I want to ask you about the issue of preference falsification. In the Bulgarian case, you interviewed a former employee of the information sociological center who used to internally report on polls showing 90-92% agreement with the official policy, and this person said to you, “we told the powers that be that this can’t be true, and they knew it.” You also talked about how in the GDR, there was a lack of trust on the part of survey subjects, so that would be an issue of preference falsification. Having looked at these other cases and the scholarly debate over preference falsification in China, which you also discussed, what are your feelings about preference falsification in China both in a recent historical context and today?

This is a great question. So to start off, what is preference falsification? The idea there is that individuals in authoritarian settings do not reveal their true opinions, their true feelings, their true attitudes towards the regime, because they’re afraid that if they’re critical of the regime, they may be punished for revealing this criticism publicly. Communist regime insiders know that, and they tend to be very skeptical of high levels of support as revealed by opinion polls. In fact, what they are interested in is levels of opposition, which is the exact opposite of support, because they know that the support numbers are inflated.

Now, what I have to say about opinion polling is that as levels of fear decline, the accuracy of opinion polls improves. We have historical evidence of this happening in Eastern Europe, in the case of the Soviet Union, but also East Germany, Bulgaria, and Poland. The quality of opinion polls improved, and the incidence of preference falsification declined as these regimes moved into the 1980s and especially the second half of the 1980s as Gorbachev’s glasnost and perestroika took on.

Of course, these are historical examples, and the question for all of us is, what is happening in China today? I’m not a complete pessimist about the accuracy of opinion polling in China– though I do think that when we get 95% of the population approving of the central government, we do have to be skeptical of that finding. However, in the last 10 to 15 years, certain types of questions which are not about support for the central government but about support for lower-level governments indicate that this level of declared support for lower-level governance in China is gradually declining. In Chapter 8 of my book, I indicate that there is a high levels of suspicion about the value of grassroots elections in China, for instance. So China is moving towards a place where the Soviet Union and Eastern Europe found themselves in the 1980s where the accuracy improved– but fundamentally what we don’t know is what percentage of people engage in preference falsification.

I also wanted to ask about the methods of the voluntary provision of information to the regime that you talked about in the book. How reliable are those methods as indicators of what the population thinks in a statistically meaningful sense? I think of things like rumors and complaints that get reported internally.

Yeah, this is a very good question. I think just like communist regimes historically developed this distrust of opinion polling and being told that 92% or 95% of the population supported them, they were not particularly interested in random sampling and statistical significance because they felt that random sampling could not be achieved. They were still interested, however, in the representativeness of the information. That is close to something being statistically significant, but of course representativeness is a broader concept.

In the case of rumors, regimes gradually became aware that the atypical one-off rumor is not particularly helpful for autocrats, because they didn’t know what to do with it. Instead, they became interested in other methods beyond random sampling for getting information which is representative or typical of a phenomenon. And in the case of complaints, for instance, and the voluntary transmission of information, the complainants are of course not randomly selected, but when these complaints get aggregated, it is also possible for the regime to detect trends. The trends focus on welfare matters. There is a very small proportion of the complaints which are about what the regime defined as political matters, and they’re heavily critical of the party, but most of them are about people wanting jobs, they want housing, in the context of China they are concerned about being displaced from their land, land requisitions and the like. In the understanding of communist regime insiders, if the material demands of the population are kept at bay, if the regime is able to deliver what people want in the material sphere, they would be less likely to engage in political opposition to the regime, so these two things are linked.

One of the very interesting counterintuitive points in the book is that complaints are something that these regimes actually want to see because they can serve as an index of people’s satisfaction and trust in the government. Your book mentioned an anecdote from Bulgaria where a regional administrator was observing that there were very few complaints, and he said, well, either we’ve solved all their problems, or they don’t trust us anymore. You also showed this with quotations from China and these other regimes to make the point that this is a systemic phenomenon.

Yeah, this is not something I expected to find before I started researching this topic in the archives. This emerged from the archives, and it emerged initially in the case of Bulgaria, but then as you point out, I do have evidence from China as well, including Xi Jinping’s father, who was quoted on this matter– he has a logic as to why more complaints are better. And then of course the tradeoff for regime insiders is that if people complain less and there are indicators that they’re unhappy, fewer complaints would then lead to more protests. From the point of view of regime insiders, a complaint is something that is done privately between a citizen and the government, and it allows for this grievance to be resolved without it being aired in the open, but it’s very hard to suppress information about protests because those are a form of contention that is presented openly and then others might join in, and then the situation can rapidly get out of control. So this relationship between complaints and protests is also something that emerged from the archival evidence I collected.

Can you talk to me a little more about the end of single-party rule in Bulgaria? You detailed how the Communist Party transitioned from having a monopoly on power to becoming a “socialist” party that has since gone in and out of the government in a multiparty system. Why did the communist party surrender its monopoly on power? You mentioned a famous quote by Mladenov, who remarked to his defense minister during a large-scale protest in December of 1989 that “it would be better if the tanks would come,” and that destroyed his political career, but why didn’t the tanks come? This was 1989; we saw what happened in China. How were things systemically different for them at that time?

Regarding what happened in 1989, for me, the process is actually quite similar in different communist regimes, which are oftentimes thought to be vastly different in the East European context. Even in the Chinese context, what we have is first a decline in citizen complaints, which regimes interpret as growing dissatisfaction, and there are objective reasons for the dissatisfaction: the economy starts to be experiencing difficulties, there are uncertainties, there are extraordinarily similar laws in both the Chinese and the East European context that allow for enterprise bankruptcy and unemployment before 1989. These are concepts that create extraordinary anxiety in urban workers, because the whole idea that you would be unemployed is something that they’re not really used to. And then we have the moves of the insiders, and in the Chinese case, we obviously have the protests, and we know how those end, with repression. In the East European context, the insiders observe what is happening in China, because the palace coups in Eastern Europe are planned in the summer of 1989 following the Tiananmen square protests, and they make a calculation that it would be very difficult to stay in power through the deployment of force, because the level of fear in the population is relatively low. So the soft-liners decided that what they would do is they would engineer changes that would not be radical changes.

Now that we have hindsight, it is hard for us to remember, or we don’t want to remember, how uncertain the situation of 1989 was. On the same day that the Tiananmen square protests were suppressed, we had the first round of competitive elections in Poland, and Solidarity eventually won a majority in one of the chambers of the Polish parliament and got a lot of seats in the other chamber. A coalition government was formed in the fall of 1989 in Poland, and Poland was the only case where the regime was undergoing a transition, but this transition was very slow. In the other Eastern European regimes, the idea was that we would allow soft-line communists to stay in power in order to prevent the opposition from taking power. And this gamble was effective for different periods of time in different regimes.

As to Mladenov and why the tanks didn’t come, I think it’s a good question. We don’t exactly have rock-solid evidence on that, but based on the argument in my book, I think the idea is that the people driving the tanks may not have obeyed at this point, because this is already more than a month after the initial opening of the regime, and there was a lot of democratic ferment, a lot of open dissatisfaction with the regime. Second, had the tanks come, this may have only led to more protest and mobilized more people to go out into the street.

If we remember what happened in Romania, where the tanks did come, that actually just happened about 10 days after the Bulgarian case. The tanks came, but that did not stop Ceacescu from being deposed, and then there was this transitional government that, although it was dominated by communists, was a softer government. But in the case of Bulgaria, Zhivkov had already been deposed, so the two cases are not exactly comparable. And actually, I think the Chinese case is not fully relevant either, because we did not have a similar development. So here the question is, who was actually in charge in China in 1989, and I guess most people would say it was Deng Xiaoping even though he did not have any formal positions beyond that of Chairman of the Central Military Commission. So he did replace individuals, Zhao Ziyang was replaced, but the big man in charge was not replaced, so the equivalent situation did not really unfold in Bulgaria. Because Zhivkov was replaced, then we have somebody who has a lot less power, a lot less charisma, a lot less control, frankly, over the political situation, and this remark, that it might be better for the tanks to come– they could not have come, because he’s a new leader whose capacity to control the armed forces is limited. And then fundamentally the levels of fear were too low for repression to have been effective.

I want to ask about what you said in the book about how the economic opportunities available to the people in power in Bulgaria eased their decision to step down. In China, which is obviously in a radically different place economically than it was in 1989, do you think that the great wealth available to the people in power is something that erodes the willingness to fight for a political monopoly?

In answering that question, one would have to first define the parameters of the parallels between the two cases. To go back to the Bulgarian case, the communist regime insiders eventually relinquished the last vestiges of control over the political system because they could monetize their membership in state security and communist party networks from before 1989. But that decision came towards the end of this transition period, and the transition period began in 1989, but it did not conclude until 1991. Giving up political power in exchange for economic power is not a good bargain for communist regime elites because they’re better off if they stay in power. This is not something that communist regime insiders do willingly; it’s a last resort.

What the Bulgarian communist party was hoping for in 1989 when it allowed for these economic reforms to start– in 1989, before this coup started to be plotted, Bulgaria legalized private firms, and this was a huge step in the right direction, moving toward a market economy. It legalized private banking. There were all these reforms that only increased the anxiety of individual citizens, but the ones at the top, the communist party insiders, knew how they could use these reforms in order to enrich themselves. But of course, they wanted to be rich and be in power.

That also reminds me of a point that I found fascinating and also somewhat counterintuitive. In Bulgaria, you talked about how the welfare state, the more highly redistributive model, became a liability for the regime, because it created a possible source of discontent. I guess intuitively, most people would think of that redistribution as something that gives the government support, that props it up and helps it, but you actually make the point that that creates risks.

Yeah, it creates expectations, and then the government feels that it has to keep delivering in order to meet the expectations. I argue that in the Chinese case, there is some redistribution, but compared to Eastern Europe, there’s a lot less redistribution– certainly under central planning there was a lot less, and now that we have a market economy, there’s considerably less. And I use welfare and redistribution as synonyms; they’re not exactly the same, but they’re in the same realm, in terms of government policies that benefit ordinary citizens. So welfare spending in China is significantly lower than it was in Eastern Europe during the communist period, and redistributive policies are less deep. However, the equivalent of the Chinese case of course is this expectation of ongoing economic growth, that prosperity will keep increasing, so when the government is unable to deliver on those expectations, this creates a moment of profound disillusionment in the population. We of course see that in China right now, and it translates into significant uncertainty for the communist party and for regime insiders. So managing those expectations of ongoing improvements in standard of living, whether they are the result of a successfully developing market economy in the Chinese case or of a constantly expanding welfare state in the East European case, that’s a difficult game for communist regimes.

The last thing I want to ask is what projects you’re working on next.

I’m working on a short book on the adaptability of the CCP, where I’m looking at four different adaptations. This is building on my book Why Communism Did Not Collapse, where I had looked at certain adaptations there, but in a broader sense across communist regimes, and this short book is strictly focused on China. It’s the first time in my entire career where I’m not doing China in a comparative perspective. I also have a China-Cuba project where I’m comparing these two revolutionary regimes and their trajectories after the revolution, post-revolutionary governance, but that is not completed.


bottom of page