North Korea is one of the most ideologized countries in the world. Its indoctrination efforts have achieved an immense ideological penetration into North Korean society. Consider the death of the long-term leader of North Korea, Kim Il Sung, on July 8, 1994. Even today, he is cherished as the “father of the nation.” When reviewing the pictures and footage from his funeral, it looks almost bizarre to see how many people poured onto the streets, desperately crying as if they had lost a close relative. Now, one obvious interpretation of this propaganda material is that the funeral—like so many other events in North Korea—was carefully staged to communicate to the outside world, but also to the inside audience, the deep and emotional bond that the suryong (great leader) had created with his fellow North Korean citizens. However, another interpretation would put forward that this spontaneous outbreak of mass grievance was not staged but expressed the genuine and authentic sorrow of the North Korean people.Footnote 1
The truth might lie in the middle. But what is beyond doubt and what needs to be taken from this historical episode is the immense societal ideologization of North Korea, whose scope has been almost unparalleled in modern history (Dukalskis Reference Dukalskis2017; Dukalskis and Lee Reference Dukalskis and Lee2020; Haggard and Noland Reference Haggard and Noland2011b; Hunter Reference Hunter1999; I. J. Kim Reference Kim1975; Lankov Reference Lankov2013; Oh and Hassig Reference Oh and Hassig2000). This paper asks the following question: how can we explain this phenomenon? How could the North Korean version of communism, the Juche ideology, spread across the whole country? In so doing, the paper concentrates on the first decades of the North Korean regime, from the 1950s onward.
The paper is thus interested in the macrosocietal relationship between indoctrination efforts on the one hand, and the creation of a society-wide belief in an ideology on the other. This relationship is not as straightforward as one might think. It is often assumed that the higher the capacities of autocratic authorities to inculcate an ideological belief into the hearts and minds of the subordinate citizens, the more an autocratic ideology is shared in a given society. The extent of an ideological belief is thus explained as a direct and immediate function of its indoctrination capacities. An autocratic ideology is spread by “drumming” its content into the people. To be clear, these top–down approaches in which autocratic regimes actively indoctrinate and intimidate their citizens explain a large part of the phenomenon (Fahy Reference Fahy2019; Hawk Reference Hawk2005). To deny this would be naïve. The paper therefore does not question these approaches. Instead, it attempts to complement the dominant top–down approach by relying on the social theory of James Coleman.
The “Coleman model” provides a microfoundation to macrophenomena, asking for individual motivations, desires, and beliefs in daily interactions. It points us to the subtler and less visible mechanisms that drive the proliferation of ideas. Even more importantly, the Coleman model also addresses the problem of micro–macro linkage and the challenge of how microinteractions accumulate to a macrophenomenon (Coleman Reference Coleman, Alexander, Giesen, Münch and Smelser1987; Reference Coleman1990; Ylikoski Reference Ylikoski and Manzo2021). It is particularly this micro–macro link that is at the heart of this paper. It will be put forward that instead of merely summarizing singular individual beliefs, we should draw analogies from epidemiological models. Like the spread of a virus, the spread of an idea can be modelled accordingly (Bettencourt et al. Reference Bettencourt, Cintrón-Arias, Kaiser and Castillo-Chávez2006; Brauer, Castillo-Chavez, and Feng Reference Brauer, Castillo-Chavez and Feng2019; Clayton and Hills Reference Clayton and Hills1993). It will be argued that—despite important caveats—the essential ideas of epidemiology can be fruitfully applied to the spread of an ideology. Three classes of parameters turn out to be of specific importance: timing, contact structure, and contagiousness—that is, the extent to which an ideology can stick in the minds and hearts of citizens.
The paper will apply these theoretical considerations to the empirical case of North Korea from the 1950s to the early 1970s. These two decades are the most striking in North Korea’s ideological development and laid the ground for an extremely ideologized society that persists to this day. Within these two decades, the North Korean Juche ideology was started from scratch and spread at an almost unparalleled speed throughout the whole country, resulting in a “Jucheization” (Park Reference Park1982, 552) of the entire North Korean society. In modern political methodology parlance, North Korea is an extreme case of rapid societal ideological penetration (Gerring Reference Gerring, Box-Steffensmeier, Brady and Collier2008a; Reference Gerring2017; Seawright and Gerring Reference Seawright and Gerring2008). I do not claim this case is representative; rather, its major strength lies in providing a good entry point into the topic. As it is so extreme, it enables researchers to adopt a “pure” perspective, with less distracting “noise.” It is at the extremes, at the exaggerations, where we see a phenomenon with the most clarity. The paper adopts this methodological perspective. As such, it allows a transfer—in lighter gradations—to other autocratic countries around the globe.
The study of North Korea still suffers from an enormous data shortage. Although survey research has made progress in the last years, particularly due to the increasing number of North Korean refugees, the country still poses severe challenges to social scientists (Chon et al. Reference Chon, Huh, Philo and Bae2007; Dukalskis Reference Dukalskis2017; Goldring and Ward Reference Goldring and Ward2024; Haggard and Noland Reference Haggard and Noland2011a; Reference Haggard and Noland2017). I base my study on secondary literature and enrich it with primary sources. I use archival documents from the former embassy of the German Democratic Republic (GDR) in Pyongyang, which are declassified for documents originating before 1980 and are accessible at the Federal Foreign Office in Berlin. The text corpus comprises around 1,200 pages.Footnote 2 It includes letter exchanges between Pyongyang and Berlin, (German translations of Korean) agency announcements and reports, summaries of diplomatic conversations that took place behind closed doors, and descriptions of daily life, as well as accounts and testimonies of official events like party meetings and demonstrations. The archival materials are of exceptional value as they offer a rare, direct, and unmediated on-site account by close political observers at the time.Footnote 3
The paper proceeds as follows. It will first outline the explanatory routes to account for the spread of an ideology in a given society. It will do so by dissecting the different causal links between the macro and the micro level. It will then turn to the empirical case of North Korea and illustrate the working of each linking mechanism with archival material and secondary literature. The paper concludes by outlining future research avenues.
Adopting the Coleman Model
How do autocratic regimes create an ideological belief among their “ordinary” citizens? The most intuitive answer is clear. It is a concert of various coercive measures, ranging from early childhood and school education, to a propaganda apparatus, and to the suppression of dissent more generally. These are features that we all routinely associate with autocratic rule and which play an integral part in enforcing the overall repressive environment of autocracies (Blaydes Reference Blaydes2018; Davenport Reference Davenport2007; Escribà-Folch Reference Escribà-Folch2013). Autocratic rulers try to indoctrinate their citizens and control their political thinking as they fear their protest. They do all they can to suppress deviant beliefs and to homogenize public opinion, resulting at best in an unchallenged and uncontested claim to legitimate rule (Gerschewski Reference Gerschewski2018). This top–down perspective is a relatively well-researched terrain in historical and comparative studies of autocratic rule (Brandenberger Reference Brandenberger2012; Brzezinski Reference Brzezinski1967; Hirsch Reference Hirsch1989; Tucker Reference Tucker1963). The causal chain starts with various indoctrination efforts that manipulate individual beliefs, which in turn foster a belief in the legitimacy of the ideology. As such, it is a top–down mechanism whose effectiveness stands and falls on the regime’s capacity to indoctrinate. Figure 1 below shows the top–down approach.
Figure 1 Top–Down Perspective: Previous Explanations of Indoctrination
The top–down approach that I term the “capacity approach” will be complemented in this article with an often-overlooked perspective. In his classic Foundations of Social Theory (Reference Coleman1990), James Coleman proposed an immensely productive “tool for social scientific theorizing” (Ylikoski Reference Ylikoski and Manzo2021, 50). Widely known as “Coleman’s boat” due to its shape, it offers an explanatory skeleton by explicitly linking the macro level of societal phenomena with the micro level of individual actions (Coleman Reference Coleman, Alexander, Giesen, Münch and Smelser1987; Reference Coleman1990, 1–26).
It is helpful to disentangle the different components of the Coleman model—that is, the nodes and the causal arrows between them (Ylikoski Reference Ylikoski and Manzo2021). In figure 2, the nodes A and D are respectively the explanans and explanandum, both on the macro level, while the nodes B and C refer to the interaction on the micro level. The famous example of Max Weber’s Protestant ethics illustrates the rationale (Coleman Reference Coleman1990, 6–9). Weber explains the emergence of a societal capitalist spirit (explanandum, node D) by arguing that a Protestant religious doctrine (explanans, node A) generates certain values on the micro level (node B) that lead to certain individual economic orientations (node C). The arrows, in turn, represent different causal mechanisms that should be analyzed separately. There is a rich strand of literature that has discussed causal mechanisms in general and Coleman’s model in particular (Beach and Pedersen Reference Beach and Pedersen2016; Bennett Reference Bennett, Brady and Collier2010; Elster Reference Elster2007; Gerring Reference Gerring2008b; Hedström and Swedberg Reference Hedström and Swedberg1998; Hedström and Ylikoski Reference Hedström and Ylikoski2010; Rohlfing Reference Rohlfing2012). For Coleman (Reference Coleman1990, 2), the “principal task of the social sciences lies in the explanation of social phenomena”—that is, the macro–macro connection, and “not the behavior of single individuals.” This macro–macro link is explained via three partial mechanisms. First, the macro–micro link explains how societal structures constrain individual actions and beliefs. For Tilly (Reference Tilly2001, 24), these “environmental mechanisms” are “externally generated influences on conditions affecting social life.” Second, the micro–micro interaction focuses on interactions on the individual level. Tilly speaks in this context of a “cognitive mechanism” that “operate[s] through alterations of individual and collective perception.” Third, the macro–micro link seeks to aggregate individual interactions back to the societal macro level—the “relational mechanism” in Tilly’s words (24). Accordingly, Hedström and Swedberg (Reference Hedström and Swedberg1996) offer a typology of the full causal mechanism and refer to “situational” (macro–micro), “action-formation” (micro–micro), and “transformational” (micro–macro) mechanisms that connect the macrolevel association (Hedström Reference Hedström, Hedström and Swedberg1998).
Figure 2 Coleman’s Explanatory Model
Against this backdrop, I share an important concern that Renate Mayntz has put forward. She argues that the third mechanism, the micro–macro link, receives too little attention in current research. She observes a “built-in bias in favor of individual action” (Mayntz Reference Mayntz2004, 250). Macrostructural determinants are only used to explain, embed, and contextualize individual action (as in the “situational” or “environmental” mechanism), but less scholarly focus is put on the “transformational” or “relational” mechanism. With her criticism, Mayntz picks up an observation that Coleman had already formulated. He put forward that it is indeed very rare that a societal phenomenon (node D) can be derived from a mere summation of individual behavior. Instead, if “the theoretical problem is, however, a problem involving the functioning of a social system … then it should be obvious that the appropriate transition cannot involve the simple aggregation of individual behavior” (Coleman Reference Coleman, Alexander, Giesen, Münch and Smelser1987, 157; emphasis in original). What is needed—and what constitutes for Coleman the proposition “of most interest” and where “Weber’s theory is the weakest” (155)—is a closer look at the mechanism that links the microlevel interactions back to the societal level. Macro-outcomes are not only pushed by microinteractions that are summed up, but there is also a macrostructural pull that should not be overlooked. Mayntz (Reference Mayntz2004, 249–52) uses—almost in passing—the intuitive example of how a technological innovation diffuses or how a rumor circulates. I take up this suggestion by Mayntz and submit that her intuitive examples demand more scholarly attention. It has been shown that the spread of ideas (and hence of ideologies) can be fruitfully assessed with the help of epidemiological models (Brauer, Castillo-Chavez, and Feng Reference Brauer, Castillo-Chavez and Feng2019; Dye and Gay Reference Dye and Gay2003). The Adding the Micro–Macro Perspective section below provides a detailed discussion about the transfer of epidemiology to political science.
To sum up, the most important and extensively studied causal mechanism is the top–down macro–micro link of indoctrination. Autocratic regimes differ in their capacity to do this. Yet while this approach is without doubt of crucial importance, it overshadows the micro–micro interactions and relegates the micro–macro link to a mere summative aggregation of individual positions. Mayntz’s analogy about the spread of rumors will be used to draw a more systematic analogy based on epidemiological models.
Explaining the Creation of a Societal Belief
In the following, I apply Coleman’s model to explain why an ideology spreads and gains more ideological followers. An ideology is defined here in neutral terms as a “cognitive structure with legitimizing functions” (Stråth Reference Stråth2006, 23; see also Gerschewski Reference Gerschewski and Wolf2025). To explain its dissemination, I distinguish between the three classes of link: macro–micro, micro–micro, and micro–macro. These three causal mechanisms are treated separately for analytical reasons.Footnote 4 In empirical reality, the dividing line between these three mechanisms is usually not as sharp as depicted here. Instead, the processes often overlap and mutually reinforce each other.
The Macro–Micro Perspective: Indoctrinating an Ideological Belief
The macro–micro link is the most researched part of the process of ideological spread.Footnote 5 I propose to subsume the various coercive attempts to make citizens believe in an autocratic ideology under the concept of indoctrination. While indoctrination has etymological roots in the teaching of a doctrine by repetition, its pejorative use has become commonplace (Callan and Arena Reference Callan, Arena and Siegel2009; Neundorf et al. Reference Neundorf, Nazrullaeva, Northmore-Ball, Tertytchnaya and Kim2024). I define it here as the inculcation of a political belief and the simultaneous suppression of alternatives, “whether through propaganda, education, or mass culture” (Brandenberger Reference Brandenberger2012, 2).Footnote 6 The term “indoctrination” thus provides an adequate way to bundle the various measures that are adopted by an autocratic regime to forcefully install a doctrine into society.
Controlling the means of education is a key element in the indoctrination of citizens. Autocratic socialization starts at an early stage, and various studies have shown that ideologized autocracies not only place heavy emphasis on controlling school curricula but also on ensuring that these measures are successful in shaping political values (Diwan and Vartanova Reference Diwan and Vartanova2020). Cantoni and colleagues (Reference Cantoni, Chen, Yang, Yuchtman and Zhang2017) have shown the causal effect of a school curriculum change in China between 2004 and 2010, which resulted in more positive views on Chinese governance and growing skepticism toward democracy and free markets. Historical scholarship has found empirical evidence of the same effect, a shaping of political values. For example, Brandenberger (Reference Brandenberger2012) and Remington (Reference Remington1988) have analyzed the impact of school education in the Soviet Union, while Rodden (Reference Rodden2010) provides an insightful sourcebook of school indoctrination efforts in the former GDR. The most emblematic example remains, however, the comprehensive efforts of the Nazi regime to “reproduce their desired vision of reality” (Hirsch Reference Hirsch1989, 63). While Adena and colleagues (Reference Adena, Enikolopov, Petrova, Santarosa and Zhuravskaya2015) demonstrate the impact of radio on the rise of the Nazi movement, the scope of Nazi societal penetration via textbooks was perhaps unparalleled in modern history (Pine Reference Pine1996). It still has repercussions in terms of anti-Semitic beliefs today (Voigtländer and Voth Reference Voigtländer and Voth2015). While indoctrination efforts do not stop at school education and often include higher education and/or further political education, the school is of particular importance not only because of the extraordinary asymmetry in authority between teachers and pupils, but also because of the timing of indoctrination efforts at school (Wojdon Reference Wojdon2017). In cognitive psychology, the years of early adulthood and late adolescence are often termed “formative years” or “impressionable years,” during which a person’s core political predispositions are shaped and remain by-and-large unchanged throughout their life (Sears and Funk Reference Sears and Funk1999).
In their classic book on totalitarian regimes, Friedrich and Brzezinski (Reference Friedrich and Brzezinski1956) argue that one of the core features of totalitarian regimes is their monopolization of the means of mass communication. This argument holds true today, as the media landscape becomes increasingly hybridized. Controlling broadcast stations, television, and newspapers, as well as social media, remains an important element in the tool kit of autocrats. Autocrats censor deviating opinions and repress dissenting voices. Again, empirical evidence from diverse places demonstrates that such control over mass media is crucial for shaping the minds of the people (Reuter and Szakonyi Reference Reuter and Szakonyi2015), dampening protest (Huang and Cruz Reference Huang and Cruz2021; Weidmann and Rød Reference Weidmann and Rød2019), distracting social awareness, and signaling the regime’s strength in maintaining social control (King, Pan, and Roberts Reference King, Pan and Roberts2017).
The Micro–Micro Perspective: Arendtian Loneliness and the Pressure to Conform
While the capacity approach focuses on the top–down perspective, the micro–micro link zooms in on the interaction between actors. The most obvious micro–micro interaction is the persuasion of one actor by another. Yet there is a more insidious mechanism at play when thinking about the dissemination of a political ideology. The classic works of Hannah Arendt provide an important perspective to help us understand the nature of this mechanism. In her sociophilosophical work on the origins and elements of totalitarian rule, she identifies in admirable clarity the anti-political principle by which these regimes operate. Totalitarian regimes are characterized by their implanting of mutual mistrust between the people, thereby destroying the essence of what she calls the “political.” For Arendt ([1993] Reference Arendt and Ludz2007), the political is generally not to be seen within the person (as the Aristotelian zoon politikon would suggest) but rather emerges in a person’s interaction with others. The political is the space between people that enables them to interact. This space is deleted in autocratic regimes when people mistrust each other. People become atomized and lose connection to their fellow citizens.Footnote 7
To fully account for Arendt’s sociophilosophical theory, it is important to mark the difference between solitude and loneliness. Recently, in a concerted effort, a group of international scholars has issued for the first time a critical edition of Arendt’s oeuvre. It compares the different versions and modifications of her classic work on the elements and origins of totalitarianism. In her “thought diary,” written at the New School in New York City between March and April 1953, Arendt argues that totalitarian regimes make people lonely, which should not be conflated with solitude. For her, solitude is defined as “being by myself, together with myself, and with everybody.” She continues to argue that “in solitude I find myself, in loneliness I lose it.” Loneliness is a state of social isolation, of being left alone, while for solitude she points to Cato: “[N]ever less solus than when solus, never more active than when doing nothing” (Arendt [1952–54] Reference Arendt, Hahn and McFarland2018, 127; emphasis in original).Footnote 8 In this sense, “totalitarianism takes care that loneliness remains and that solitude cannot develop. You can be lonely among many people, but never solitary. Therefore, they do not leave you alone. … They do not permit to isolate yourself” (128).
By keeping people lonely and emotionally disconnected from their fellow citizens, these regimes disable normal interactions between ordinary citizens. Such a societal climate lets autocratic ideological beliefs mushroom and diffuse. Lisa Wedeen (Reference Wedeen1999) shows for the Syrian case that by inundating public and private life with the symbols of the regime, the personality cult surrounding Hafiz al-Assad made people tired and worn out, isolated citizens from each other, implanted mutual mistrust, and defined the parameters of what could be said and done in public. It resulted in pro-regime actions even when people did not share an ideological belief in the regime. Myanmar is another case in point. Here, Hlaing (Reference Hlaing2009, 277) reports that the “golden rule” is that “nobody wanted to be the troublemaker,” and citizens stayed out of the private domains of others to avoid any conflict.
Elisabeth Noelle-Neumann (Reference Noelle-Neumann1984) termed this self-reinforcing logic the “spiral of silence.” As people fear social isolation and feel pressure to adapt, they closely observe the behavior and attitudes of their social surroundings to evaluate what the most “appropriate” behavior and attitudes are in public. People shy away from publicly speaking out against what they perceive as dominant opinions. In some circumstances it is not only passive self-censorship at work, but even active preference falsification. As we learn from Timur Kuran (Reference Kuran1997, 3), preference falsification is a specific form of lying. It is the “act of misrepresenting one’s genuine wants under perceived social pressure.” It is not lying about actions, but about the underlying motivations. The obstinacy of communism was for Kuran therefore based on active falsification of preferences (Reference Kuran1997, chap. 7).
Yet this should not overshadow the reality that—nolens volens—a smaller or larger segment of the population feels attracted to the prevailing political ideology (critical: Mercier Reference Mercier2020, 128–34). They are genuine supporters of the ideology and do not need to be persuaded. In a recent, highly acclaimed novel (that almost morphs into a history textbook), Antonio Scurati (Reference Scurati and Appel2022) vividly describes the rise of fascism in Italy. What makes this novel so compelling is the inner perspective of the early supporters of Benito Mussolini, the demoralized, disgruntled, traumatized, and disadvantaged veterans from World War I (the Arditi), and the deep emotional bond that united them. In a similar vein, the German Freikorps that initially constituted one of the organizational backbones of the Nazis were “formed on the basis of close emotional relations” (Linz Reference Linz and Linz2000, 18). It would be shortsighted to argue that only these paramilitary groups were fervent ideological supporters of their respective regimes. Instead, as historian Hans Maier (Reference Maier and Maier1996, 244) aptly characterizes, the full-fledged ideologies of communism or fascism had a “tremendum et fascinosum” that made them attractive to ordinary citizens. The inherent appeal that autocratic ideologies have for some segments of society should not be overlooked (Gerschewski Reference Gerschewski2023, chap. 2).
The micro–micro link in autocratic settings sheds light on the interactions between people. If a societal climate of mistrust and loneliness is nurtured and institutionalized, the diffusion of an ideological belief is not only the direct product of indoctrination but also a by-product of less visible social adaptation pressure. This conformity pressure is felt even more when the political context is characterized by a strict ideological orthodoxy and an already well-progressed societal isomorphy. As will be discussed in the next section in more detail, this tendency is further augmented when the ideology taps into national resentments and thus becomes more attractive in the eyes of ordinary citizens.
Adding the Micro–Macro Perspective: Learning from Epidemiological Models
I put forward that we can learn more about the micro–macro link by borrowing insights from epidemiological models. Other disciplines have already used epidemiological models sporadically. In history of science, for example, Bettencourt and colleagues (Reference Bettencourt, Cintrón-Arias, Kaiser and Castillo-Chávez2006) ask how the Feynman diagram, a model in theoretical physics that depicts subatomic particles, was increasingly adopted in the first seven years after its invention in 1948 by different academic communities in the United States, Japan, and the USSR (Kaiser Reference Kaiser2005). A second example is the work of Sperber (Reference Sperber1996, 96), who is interested in a naturalistic anthropological explanation for the dispersion of culture. Unlike “intuitive beliefs,” whose spread is determined by cognitive factors, “reflective beliefs” are spread by ecological factors. Third, sociology has taken up epidemiological models to explain the prevalence of riot behavior as well as norm cascades (Bonnasse-Gahot et al. Reference Bonnasse-Gahot, Berestycki, Depuiset, Gordon, Roché, Rodriguez and Nadal2018; Burbeck, Raine, and Stark Reference Burbeck, Raine and Stark1978; Centola, Willer, and Macy Reference Centola, Willer and Macy2005).
In this paper, I aim to draw a systematic analogy and use epidemiological models as a heuristic for better understanding the dissemination of a political ideology. While the analogy might be intuitive, three caveats are important to highlight. First, while the spread of an idea is often based on intentional action by the transmitter and/or adopter, the spread of a disease is unintentional. Second, there is no simple and straightforward mechanism by which ideas disperse. In contrast to the transmission of a virus, which usually involves a medical reaction once infected, it is more appropriate to talk about the transition probability of a new idea being adopted, given that the new idea might be met with skepticism. Third, it often proves empirically difficult to find sufficiently granular and reliable data for the spread of an idea. Yet, despite data limitations, epidemiological models can provide key parameters that are essential in understanding the spread of an idea.
At the heart of these models lies the division of agents into different classes that reflect their epidemiological status. It is common to distinguish between five types of agents: susceptibles (S; potentially the whole population), idea incubators (E; people who have been exposed to the ideology but have not yet adopted it), idea adopters (I; believers), skeptics (Z), and finally the recovered (R). The last class perhaps needs more elaboration. The recovered represent those who have been exposed to an idea, but have lost interest in it and have been “immunized” against it. For the context of this paper, the recovered are those people who once believed into the official ideological claim, but then turned their back on it.
When it comes to the most important model parameters, three clusters of factors can be identified. First, transition probability is important. On the one hand, if people come into contact with adopters, one needs to consider their transition probability from susceptibles to adopters (S–I) and susceptibles to incubators (S–E). On the other hand, if people come into contact with skeptics, their transition probability of turning into skeptics (S–Z) and incubators (S–E) is key. These interactions happen on the micro level. Second, the societal contact structure is decisive when aggregating from the micro to the macro level. The per capita contact rates between susceptibles and adopters (S–I), as much as between incubators and adopters (E–I), and between susceptibles and skeptics (S–Z) need to be accounted for. The environment in which these microinteractions take place is important. Macrofeatures such as population density, but also cultural norms of social distance, are particularly relevant. Third, temporal factors are critical: the timing of an idea, its average lifetime, and its incubation time all need to be accounted for when applying an epidemiological model to the diffusion of an idea. When translating these factors to the realm of political science, and in particular to the dispersion of a political ideology, the conceptualization of these factors cannot be seamless. Instead, they need to be adapted and translated to the empirical phenomenon under study. These three factor clusters are discussed in more detail below.
Transition Probability and Contagiousness
When it comes to transition probability—both in favor of and in opposition to a political ideology—the translation task is defined by the following question: when a susceptible person (potentially any citizen) is exposed to an adopter of an ideology, how likely is it that the susceptible turns into an adopter? Viruses differ in their degree of infectiousness, and vary in their “strength” to infect susceptibles. Some viruses jump from one person to another via the slightest droplet, others are less contagious. Translated to the study of political ideologies, the key question is how strong is a political ideology’s power to persuade citizens. What makes a political ideology so convincing that it jumps easily from adopters to susceptibles?
David Beetham (Reference Beetham1991, 17) argues that it is the “justifiability” of an ideology. A guiding ideology must address widely held societal beliefs, even emotional resentments and angers, so that the ideological claims speak directly to people’s demands. Ideologies (mis)use past experiences, such as a shared colonial history and struggles for national independence, but can also employ religious dogmas and (politicized) ethnic claims. The better a political ideology can relate to the national psyche and hark back to widespread societal sentiments, the greater its justifiability and hence its persuasive power vis-à-vis other people. Ideologies need to hit a societal nerve.
Contact Structure
The second factor refers to the contact structure. A virus diffuses more easily in groups whose members regularly come into contact with each other than in groups where contact is less frequent (e.g., densely populated urban areas compared to sparsely populated rural areas). Translated to the realm of political ideologies, autocratic regimes must make sure to mobilize people and enforce social contact. Susceptibles need to be constantly exposed to the ideological claims. Densifying space between individuals and contracting the time between interactions is central to the spread of an ideology. To increase ideological penetration, autocracies need to build societal structures that impose close, frequent, and inescapable ideological contact.
The Arendtian loneliness of isolated people and autocratic attempts to dissociate people from each other are not in conflict with this idea. What Arendt describes is a loss of mutual emotional trust between citizens, complicating genuine interactions. Coupled with the strong incentive for preference falsification and social conformity pressure, citizens in autocratic regimes cannot be sure what their counterparts think. Skeptics often do not dare to reveal themselves as skeptics. Autocrats not only create but also exploit this interpersonal uncertainty in their favor. An ideology mushrooms when people are exposed to others they assume are already adopters of the ideology. As a result, the number of ideological adopters is constantly overestimated. The latest experimental research has found that even the perception of autocratic popularity breeds actual popularity. Taking clues from assumed popularity, this type of endogenous legitimation works insidiously: if citizens assume that the autocrat is popular with the people, they assign higher popularity to him as well (Buckley et al. Reference Buckley, Marquardt, Reuter and Tertytchnaya2022). This makes the mutual mistrust between people so consequential.
Temporality
Temporal factors are essential in explaining the proliferation of a virus. To take full advantage of exponential growth, a sharp increase in adopters at the beginning of the pandemic leads to a rapid and accelerated growth of adopters. An exponential dynamic is reached when the lifetime of the virus is sufficiently long and the incubation time low. Again, translated to the study of political ideologies, susceptibles need to be overwhelmed within a very short time span. Adopters and incubators need to rise quickly at the beginning. Instead of a steady, slow, and gradual increase, political ideologies benefit from spreading rapidly at the beginning. This is why timing is so important: an ideology needs particularly favorable temporal circumstances—that is, the creation of a window of opportunity in which momentum can be exploited and society overwhelmed. The mathematical force of exponential growth can then exert its full impact.
Figure 3 provides an overview of the theoretical argument. It combines the capacity approach of top–down indoctrination with a complementary perspective that emphasizes the micro–micro interactions and the micro–macro dynamics. In the following, this theoretical reasoning will be applied to the North Korean case.
Figure 3 Explaining the Spread of an Ideology
What Is “Spread”? The North Korean Juche Ideology
North Korea’s political ideology is the Juche ideology. Interpretations of this ideology vary. It is either a complex set of ideological propositions or an enormously simplistic account that is simply “good enough” to justify power and its abuses. Juche is often equated with self-reliance, autonomy, and self-identity. A more precise translation is offered by Oh and Hassig (Reference Oh and Hassig2000, 17). In the literal sense, juche here means “subject,” with ju as “main, essential” and che as “body, self, origin.” Kim Jong Il (Reference Kim and Jong Il1985, 20), the son and successor of the state founder Kim Il Sung, described it later as the “attribute of man who is desirous of living independently as master of the world and his own destiny.” Despite the strong emphasis on human capacities, men still need guidance that is provided by the suryong, the “beloved leader.” Suryong is an honorary title given to Kim Il Sung. He led the people for the revolutionary cause and the people must obey the demands of the leader (I. J. Kim Reference Kim2003, 63–64).
The Juche ideology can best be described as an amalgam of Marxism-Leninism, traditional Confucian thoughts, and a strong sense of nationalism. It depicts itself as scientific and as having a privileged insight into the course of history. Society should be transformed holistically. The Juche ideology is propagated as a further development and perfection of Marxism-Leninism (I. S. Kim Reference Kim1975). The route to the Marxist utopia of a classless society can be reached in North Korea directly from feudalism. In North Korea, it is claimed that the “objective conditions”—that is, North Korea’s colonial past and its geostrategic location—have met the “subjective condition,” which is seen in the creative strength of the leader Kim Il Sung, who was therefore “virtually a divine gift to the people” (Oh and Hassig Reference Oh and Hassig2000, 23).
Given the later ubiquity of the Juche ideology, it is astonishing that it was first mentioned only in 1955, seven years after national independence. The term did not exist before that date. Lankov (Reference Lankov2002) shows that juche had no entry in the 1959 official dictionary of the Communist regime and only a very sparse entry three years later. In a similar vein, Suh (Reference Suh and Suh2000, 41) demonstrates that juche was not mentioned in public speeches until 1963. At the beginning, it was therefore more of a “political slogan than an ideology” (Park Reference Park1982, 548).
Explaining North Korea’s Ideological Penetration
In the following, I will explain North Korea’s deep societal penetration with the Juche ideology from the 1950s to the early 1970s, applying all three of the partial mechanisms that govern an ideology’s spread.
The Macro–Micro Perspective: North Korea’s Indoctrination Capacities
As described above, the macro–micro perspective is the most researched. It argues that the extent of ideological penetration is a direct function of the extent of indoctrination efforts. This is a clear-cut and straightforward argument. And, indeed, North Korea’s efforts in this area have been immense (Ba, Greitens, and Kim Reference Ba, Greitens and Kim2023; Fahy Reference Fahy2019; Hassig and Oh Reference Hassig and Oh2009; Hawk Reference Hawk2005; Hunter Reference Hunter1999; J. Lee Reference Lee2024; McEachern Reference McEachern2019; Oh Reference Oh1988; Park Reference Park2002). Both in terms of its propaganda and its education efforts, North Korea is almost unparalleled (M. Lee Reference Lee2024). When, at the beginning of the 1980s, a group of Western scholars was allowed to travel to the country, Koh (Reference Koh, Kim and Koh1983, 31) reported that the daily rhythm of ordinary North Korean citizens consisted of three parts: a third of the day was dedicated to work, a third to sleep, and a third to the study of ideological texts by Kim Il Sung. Archival evidence of on-site assessments at the beginning of the 1970s also described the pervasive education of all parts of the North Korean population that begins in the cradle and at kindergarten.Footnote 9 Similarly, decades later Hunter (Reference Hunter1999, 22) described the surprising accuracy with which schoolchildren could remember dates in North Korean history and on-site instructions given to them, as well as passages of ideological propaganda texts. Even before entering school, children were educated in the spirit of “Father Kim Il Sung” and attended various youth organizations and “voluntary services” before going on to perform compulsory military service and pursue a party career (30–57).
Kim himself was actively elevated to an almost godlike status and presented as infallible and omniscient. He was praised as the “sun of the nation,” and his seemingly “endless love” for the North Korean people was demonstrated in massive buildings and ceremonial halls. The monumental 170-meter-high Juche Tower that was opened in 1982 for his 70th birthday celebrates the life of the national founder with 70 stairs for each year and 25,550 steps for each day he lived (I. Lee Reference Lee1991, 141). His birthplace at Mangyongdae became a holy site of pilgrimage, as did his mausoleum following his death (Hunter Reference Hunter1999, 17; Koh Reference Koh, Kim and Koh1983, 28–30). The North Korean calendar begins with his birth year: 1912 CE is the year Juche 1. The religious concept of “eternal life” has also been adopted to serve him after his death (Park Reference Park and Park1996), further showcasing North Korea’s enormous indoctrination efforts.
Complementing the Capacity Approach, Part 1: Micro–Micro Interactions
Mechanisms of the micro–micro interaction—the pressure on individuals to conform and the institutionalization of mutual mistrust—have been outlined above. Arendt’s dictum of loneliness and her concept of an atomized society in which individual members have lost the capacity to trust each other have been highlighted. When Alexander Dukalskis conducted interviews with North Korean defectors in 2011, the atmosphere of mistrust was still omnipresent in their memories. He provides various accounts of North Korean citizens from different social strata that all refer to interpersonal mistrust: “Even within the family, discussion in North Korea is not entirely free” (Dukalskis Reference Dukalskis2017, 39). According to one interview partner, it is only among the “best-trusted friends” that political issues would be discussed within certain limits, not with “new friends” or “normal friends,” even if the bonds of friendship go back to elementary school. Others submit that they avoid talking about politics altogether, “so that discussions never really happen” (38–39). Smith (Reference Smith2015, 120) is therefore right in her assessment that “North Koreans became socialized within a culture of secrecy.” North Korea’s culture of omnipresent societal secrecy is the most favorable environment for nurturing loneliness and mistrust among North Korean citizens. It drives a spiral of silence and sets enormous incentives to falsify genuine preferences. North Koreans suffer from severe social pressure, and ideological misbelief is harshly sanctioned. The potential costs of noncompliance with the ideology are extremely high. North Korea’s apparatus of repression is one of the most punitive in the world, and includes a specialized prison camp system (Haggard and Noland Reference Haggard and Noland2009; Hawk Reference Hawk2005).
The archives of the GDR embassy in Pyongyang present an instructive example of how mutual mistrust has been subtly exploited by the regime. For the 60th birthday of Kim Il Sung in 1972, badges of different designs were distributed to all North Koreans, who were required to wear them in public. These badges were worn openly, and divided the North Korean population into three groups according to their ideological believer status. Foreigners did not receive a badge.Footnote 10 While these observational reports cannot reveal the real extent of ideological belief in North Korea (and the actual distribution of belief across society), we can draw interpretations from them. The sorting of North Koreans in this manner reveals an inclusionary and exclusionary purpose. Most obviously, it created a North Korean community without foreigners. The mandatory public wearing of the badge functioned as a general display of ideological belief among North Koreans, most probably overselling and exaggerating the degree of actual belief. Nevertheless, this in-group was then insidiously divided into three subgroups that mistrusted each other, envied each other, and/or viewed each other with suspicion, creating an exclusionary competition between citizens. As a result, mutual mistrust was sown and the Arendtian political space between people was destroyed.
North Koreans cannot be sure about the real intentions of their conversation partners. They might be genuinely interested in the political opinions of others, or they might be whistleblowers and secret regime informants. And punishments for critical political statements are draconian. This leaves people in a state of behavioral uncertainty that autocratic regimes are keen to maintain. In other words, skeptics do not reveal that they are skeptics. The signaling game of inferring from clues is just too risky, leading to a constant overestimation of the number of ideological adopters—a tendency that itself breeds actual adopters. A cultural norm of mistrust is deeply ingrained in North Korean citizens. The spiral of silence works efficiently in North Korea.
Complementing the Capacity Approach, Part 2: The Micro–Macro Link and Epidemiology
It has been argued that epidemiological models can be fruitfully used as a heuristic to better understand the spread of an ideology. Three classes of epidemiological parameters have been singled out as most important determinants of an ideology’s spread: (1) the contagiousness of the ideology via its inherent justifiability, (2) the societal contact structure, and (3) timing and temporality.
Transition Probability and Contagiousness
As depicted above, the North Korean Juche ideology was built on widely held societal views. It is deeply anchored in postcolonial struggle and hypernationalistic sentiment, a neo-Confucian understanding of leadership, and an immense personality cult. When Kim Il Sung propagated his political course in the 1950s, it struck a chord with ordinary North Koreans. The Korean War had just ended and memories of the humiliating phase of Japanese control from 1905 to 1945 was still vivid. This period is depicted as a national trauma in which the Korean nation was not the master of its own destiny, so a hypernationalistic variant of communism seemed—as it did to other postcolonial communist regimes—to be a promising route to power. The archival materials that I retrieved from the embassy of the GDR reveal that North Korea’s nationalistic course was even a topic of international complaint. Already in the early 1960s—that is, only a few years after the invention of the Juche ideology—North Korea’s fellow communist states warned in secret diplomatic meetings that the ideology deviated significantly from Marxism-Leninism.Footnote 11 In the eyes of the GDR officials, the Juche ideology represented a “nationalistic conception” that was “bourgeois-revisionist” and “eclectic”Footnote 12 and even “contradicts Marxism-Leninism.”Footnote 13 A representative of the Bulgarian embassy described the political guidelines as being characterized by “inconsequence, without principles, lack of continuity, national egoism, and pragmatism.”Footnote 14 It is therefore fair to say that the Juche ideology was, almost from its very inception, eyed with suspicion by other socialist countries for its deviance from communist orthodoxy.
North Korea’s ideological discourse was also connected to the tradition of neo-Confucianism widely held in Korean society. The country’s leader was from early on depicted as a patrimonial leader, which found its Confucian mirror image in the father–son and ruler–ruled relationships that emphasize obedience and filial duty. The traditional Confucian image of the “mandate of heaven,” in which the ruler has the moral task of taking care of the ruled, is routinely invoked by North Korean leaders. The ruler should rule in the name of the ruled and for the common good of the people who, in turn, follow his example and obey (Cornell Reference Cornell2002, 119–27). In a similar vein, hierarchical thinking is also conveyed through organicist images. In its basic understanding, this perspective depicts North Korean society as a sociopolitical body in which the leader serves as the brain that develops ideological imperatives while the political party is the nerve system that permeates the body and transmits decisions to ordinary people (Park Reference Park2002, 33–35).
From the very beginning, the Juche ideology was an attempt to blur the lines dividing the nation, the regime, and the leader. This merging was developed further when, in 1974, “Kimilsungism” (Kim-il-sung-chui) emerged as an official term in North Korean propaganda. In the archives, the GDR’s ambassador to North Korea, Kurt Schneidewind, described in 1961 that a strong personality cult already surrounded Kim Il Sung, which expanded further over the years.Footnote 15 In 1963, an “overemphasis of the role of comrade Kim” was conveyed and other important comrades like Kim Il and Park Kun-Chen were marginalized.Footnote 16 An assessment in 1970 talked about a growing “glorification”Footnote 17 of Kim and two years later, a report aptly described the “worship” directed toward Kim Il Sung, which reached “forms of fanaticism.”Footnote 18 This Kimilsungism shows that the North Korean nation found its expression in the Juche ideology, the regime readily adopted it, and the ideology became identical with the leader. Kim’s personal goals, the aims of the regime, and the raison d’etat had become congruent and a distinction between them could not be drawn.
Societal Contact Structure
In epidemiological models, the per capita contact rate is of key relevance. When it comes to the dissemination of an ideology, the densifying of interpersonal contact is crucial. This is done by compressing the space between people and contracting the intervals between which they interact with each other. While the above-described mutual mistrust among citizens prevents people from identifying skeptics and even the immunized or recovered, the densifying of societal contact structures is meant to expose all susceptibles as often as possible to (seeming) adopters. This densifying is done via two routes in North Korea. On the one hand, the number of contacts is increased. On the other, the North Korean regime impedes both upward and downward social mobility.
In North Korea, society is divided into numerous subunits (inminban). The core purpose of these subunits can be interpreted as the reduction of interpersonal distance and the increase of interpersonal contact frequency. The whole society is broken into smaller units that on average consist of 20 to 50 neighboring families (Lankov Reference Lankov1995). These groups are headed by local party officials who oversee the upholding of the moral integrity and ideological loyalty of their members. The experiences of the numerous inminbans are also used to pool information. Comparing insights, learning from best practice and bad practice, and coordinating efforts centrally has led to decreasing marginal costs with growing numbers of inminbans. They constitute a hierarchical, interwoven, dense, and coordinated network of believers that spans the whole country, artificially increasing the number of contacts between ideological believers and other citizens.
Minimizing contingency and maximizing adaptive expectations is one further ingredient in the North Korean recipe for spreading an ideological belief. Since as early as 1946, directly after gaining independence from Japan, the whole of North Korean society has been structured along the personal songbun. Songbun is a concept that clarifies personal backgrounds. With the help of the land reform in 1946, the structure of North Korean society was overturned in an ideologically motivated “far-reaching social revolution” (Scalapino and Lee Reference Scalapino and Lee1972, 350). Every citizen received a certain songbun according to their level of loyalty to Juche ideology, socioeconomic status, class, family, regional background, and degree of personal suffering from war. Workers, farmers, and military officers, particularly those who had fought in the struggle against Japanese colonialism, had a very favorable songbun and a privileged position in society. Lankov (Reference Lankov1995) argues that there were three broad classes and 51 subcategories. At the Sixth Congress of the Korean Workers’ Party in 1980, the estimates were that a quarter of North Koreans belonged to the regime-loyal core class, half to the “wavering class” that needed to be viewed with suspicion, and finally a quarter that belonged to the “hostile class” (Oh and Hassig Reference Oh and Hassig2000, 127–47).
Every citizen’s songbun is recorded in detail and the regime continues to keep track of it to the present day. One’s songbun decides one’s future personal possibilities. It determines one’s education and access to higher schools and universities, jobs, housing, the health system, and the judiciary; it also determines one’s possible marriage partner. This hypersensitivity toward personal background ensures the demobilization of North Korean society. It is a caste system, and vertical social mobility is prohibited. Although citizens do not have a detailed knowledge of their songbun, they know which broader class they are in and what options they have in life. As a result, the double movement of fixing societal status via a songbun system and of densifying contact structures via inminbans contributes significantly to the successful dissemination of the Juche ideology.
Timing and Temporality
The year 1956 is of heightened importance in North Korea. Favorable internal and external circumstances were created to allow for massive initial ideological investments. The Juche ideology pervaded the country and overwhelmed the susceptible population, making full use of its rapid initial growth.
The year was also a watershed moment for other communist countries. Like the leaders of other communist parties, the North Korean leadership faced factional struggles. The “August group,” which was inspired by Soviet leader Nikita Khrushchev’s “Secret Speech” of 1956, hastily planned a coup against Kim at the 1956 Plenum of the Korean Worker’s Party Central Committee (Lankov Reference Lankov2005; Szalontai Reference Szalontai2005). They took Khrushchev’s open critique of Stalinist rule as a template for criticizing excessive power and the nascent personality cult surrounding Kim Il Sung. They demanded collective leadership instead. Kim feared losing ground, and successfully fended off opposition demands by threatening, blackmailing, and repressing dissent.
This domestic power challenge was coupled with an international challenge: the Sino-Soviet split. Yet far from forcing North Korea to choose sides, this rift played into Kim’s hands. Kim skillfully mastered balancing the two communist powers, treating the USSR as North Korea’s economic motor and China as its cultural mentor as it steered its own nationalistic course. This is well documented in the archival material. In several documents, representatives of other socialist countries accused the North Koreans of not being in line with Soviet doctrine. Rumors were reported that North Koreans perceived Khrushchev as “a worse revisionist than Tito.”Footnote 19 North Koreans “agitated” against Khrushchev and the leadership of the Soviet communist party as well as other European socialist countries.Footnote 20 Moreover, some reports include complaints about how North Korean authorities would downplay or even ignore the role that international support, particularly from the USSR, played in building up the country and freeing it from colonial Japan.Footnote 21 North Korea was accused of following the Chinese course of “petit bourgeois nationalism, big power chauvinism, racism, adventurism, and neo-Trotzkyism.”Footnote 22
On the other hand, archival documents also reveal Kim Il Sung’s strategic criticism of the Chinese. A Soviet representative reported in 1966 that Kim Il Sung would assess the Maoist Cultural Revolution as a “huge stupidity” that would also impact his country considerably. The representative asserted that Kim Il Sung argued that North Korea would not be able to openly criticize China as it would need to think of the future, but it would be aware of the “malicious behavior” of the Chinese.Footnote 23 A Cuban representative added to this picture, reporting in 1966 that jokes could be heard in the streets of Pyongyang about the Cultural Revolution and about Mao himself, who was said to be senile and could only be cured with Korean ginseng roots,Footnote 24 while Kim Il Sung told a Hungarian delegation that, from his point of view, there would be no dispute between China and North Korea.Footnote 25
The picture that emerges from the archival material is that Kim Il Sung was indeed clever and skillful enough to use the mutual suspicion between the USSR and China for his own purposes. The Sino-Soviet split opened a window of opportunity for his nationalistic course and allowed him to cultivate his own ideology in the shadow of the two superpowers. While the two socialist countries were preoccupied with each other, the North Korean leadership could follow through with its own ideological principle of political independence.
In a favorable situation in which internal power struggles were resolved and international partners played against each other, the Kim regime invested massively in indoctrination efforts after 1956, leading to a vast exponential growth in ideology adopters and incubators. While exact data are difficult to retrieve, Yiu (Reference Yiu1969, 22) estimated that already in 1961—only six years after the first public mention of the Juche idea—there were three million copies of North Korean propaganda material in circulation. Remember that the word juche did not exist in Korean vocabulary until the end of the 1950s. In 1961, the collected works of Kim Il Sung already encompassed six volumes with five hundred pages each and a circulation of three hundred thousand for each volume (Suh Reference Suh1981). A monopoly over the means of mass communication was also quickly established. The daily newspaper Rodong Shinmun, state television, and radio programming expanded rapidly from the late 1950s and became effective propaganda instruments (Park Reference Park2002, 54–59). Furthermore, the Juche ideology not only defined the political and economic realm but also reached far into the daily life of the people. Juche music, Juche art, Juche dance, Juche literature, Juche architecture, and Juche design are among the more bizarre examples of Juche’s penetration into the cultural sphere, and were already widespread in the early 1970s.Footnote 26 The speed with which the information monopoly was erected and the enormous exponential growth of ideological adopters and incubators from the moment the Juche ideology was launched helps to explain why it was so successful. Like the spread of a virus, the Juche ideology pervaded North Korean society at an almost unparalleled pace, creating a deep-seated ideological belief that has persisted for many decades.
Conclusion
The major motivation for writing this paper has been to shed more light on the spread of an autocratic ideology. I took North Korea as an extreme case in which the working mechanisms of autocratic ideology can be demonstrated with the greatest clarity. North Korea created a pervasive societal belief in the Juche ideology in such a fast and sweeping move that it might not be too far-fetched to argue that it constitutes the most ideologized regime worldwide.
The capacity approach—that is, top–down indoctrination efforts—is the dominant approach to explaining this phenomenon. Communist education, media propaganda, and large-scale repression of deviant behavior explain a good deal of why ideological penetration has been so successful in North Korea. This holds true for other communist and autocratic regimes, and this paper does not question the utility and importance of this approach. Nevertheless, the paper addresses two shortcomings. First, we tend to turn a blind eye to micro–micro interactions. And, second, as Coleman lamented, we often refer only to a summative aggregation from the micro level to the macro level, neglecting societal dynamics that are driven by institutional and structural features. In this situation, epidemiological models have served as a useful heuristic to help us think more thoroughly about the micro–macro link.
Having a closer look at these neglected links does not mean that the spread of an ideology comes about automatically and is nonagentic. On the contrary, it has been demonstrated that, for example, the densifying of societal contacts or the massive initial ideological investments needed to spark the exponential growth of an ideology were carried out by actors in the North Korean regime. The strict songbun system and the concept of societal subunits (inminbans) did not fall from heaven but were created as much as the monopoly of communication was enforced, TV stations established, newspapers founded, and propaganda publications printed. Yet it is important that explanations for the spread of an ideology unfold a dynamic that goes beyond the top–down approach. A micro–micro spiral of silence, the widespread creation of loneliness and mutual mistrust, an increase in the contagiousness of an ideology brought about by providing justifiable ideologemes, and a densification of contact structures between individuals have their own inbuilt structural effect. I submit that these inner dynamics should not be neglected. Instead, it has proven to be fruitful to look across disciplines and learn for future research from epidemiology, be it as a heuristic, as guidance for qualitative field notes, or, if the granularity of data allow, for formal models. The study of how an ideology, technology, knowledge, or any other diffusible item spreads might benefit from such a cross-fertilization.
Acknowledgments
I would like to thank the four anonymous reviewers as well as the editors of Perspectives on Politics for their stimulating comments and helpful advice. The paper also benefited from insightful discussions at the “Varieties of Communism” conference at the University of Florida, Gainesville, in February 2023. I am particularly grateful to Michael Bernhard, Iza Ding, Jeffrey Kopstein, and Dan Slater. For critical feedback, I would also like to thank Nora Wacker. The open-access publication was funded by the WZB Berlin Social Science Center.
Open access
