■ See Korean Version on EAI Website

1. Introduction: An Era of Rupture, Yet the Inflection Point Will Come Again

At the end of 2023, North Korea defined the South and the North as ‘two hostile states’ and erased the very word ‘unification.’ The channels for dialogue have closed, and exchange has effectively ceased. At present, little to no connection remains between the two Koreas. However, this diplomatic freeze is unlikely to be permanent. Over the past seven decades, inter-Korean relations have always been a pendulum swinging between rupture and thaw, and the inflection points of transition have always arrived without warning. The real question is how prepared we are when that moment arrives. And the core of that preparedness lies in the ability to read North Korea’s changes first, and accurately.

How, then, have scholars traditionally analyzed North Korea? For the most part, we have interpreted it on the basis of the ‘Supreme Leader’s instructions’ or the ‘Party’s official documents.’ This study challenges this conventional practice directly. If one reads only North Korea’s policy announcements, every decision appears to emerge from the mind of a single supreme leader. Thus, we have long explained North Korea’s policy changes as the unilateral decisions of the Supreme Leader. This is the so-called ‘Supreme Leader determinism.’

However, as research on authoritarian regimes consistently points out, even a dictator cannot rule alone. When the direction of policy changes, the elite group that had benefited from the existing line is pushed aside, and a group to support the new line rises. The grievances of the group that suffers losses soon become a threat to the leader. Therefore, a major policy shift is likely to be the product of a strategic power struggle between the leader and the elite.

The question of this study is straightforward. Is a major shift in state strategy during the Kim Jong Un era the unilateral decision of the Supreme Leader, or is it the result of a specific elite group rising first and driving it? Policy may not begin only at the moment it is announced. The composition of the Supreme Leader's inner circle may shift prior to a formal policy announcement. If so, by tracking that change, we can read the direction of North Korean policy in advance. This study seeks to answer this question with empirical evidence.

This work also contains an attempt at the methodological level. Because of the regime’s opacity, North Korean studies have long relied primarily on discourse analysis and interpretation of anecdotal cases. However, if we can systematically collect and verify the various signals that leak out through the cracks of the regime—just as we accumulate and analyze data on the figures around the Supreme Leader—we will be able to make more accurate and scientific estimations. The title of this study, 'Evidence-Based North Korea Studies,' explicitly reflects this methodological approach. However, this presentation does not claim to complete that approach but merely proposes a direction. Based on data on the elites accompanying Kim Jong Un’s public activities, it is a kind of exploratory study demonstrating how such a methodological shift might be made possible.

This study proceeds in the following order. Chapter 2 examines why the research question of this study matters in relation to prior research; Chapter 3 introduces the theoretical lenses of power-sharing and ‘Red and Expert.’ Chapter 4 presents the data and methodology; Chapter 5 presents three competing hypotheses; and Chapter 6 shows the four shifts in strategic line through ‘four scenes.’ Chapters 7 and 8 present the study’s academic implications and policy recommendations, respectively. The final Chapter 9 addresses the limitations of this study and directions for future research.

2. Why This Study Matters: Its Place within Prior Research

Research on authoritarian regimes has long presupposed ‘leader-elite interaction’ as the core mechanism of politics. This is because, in regimes with weak institutional legitimacy such as elections, the leader must employ various strategies to co-opt or repress elites in order to survive. Yet the great majority of empirical research still focuses on the survival strategies of the ‘leader.’ While explaining how coups and popular uprisings are managed, and why institutions such as parliaments and parties are operated, elites have been perceived not as active agents but as passive agents manipulated by the leader’s strategy.

The discussion of the ‘winning coalition–selectorate’ in Bueno de Mesquita et al. (2003), and the discussion of institutions and cooptation in Geddes et al. (2018), also share a leader-centered perspective in that they ultimately ask ‘which elite-management strategy is advantageous to the leader’s survival.’ In other words, elites were an object of analysis, but not a subject of action.

2.1. A Recent Shift: Elites as ‘Agents’

Recent research, recognizing this limitation, has shifted the focus of analysis from the single leader to the background, composition, and influence of elites. Newson & Trebbi (2018) analyze the activity records of elites in North Korea, China, and Africa, and argue that the distribution of authoritarian power operates not as ‘winner-take-all’ but as ‘winner-take-more.’ That is, while the leader takes more, a considerable degree of power is also distributed to internal factions and elites. Shih & Lee (2023) show, using Chinese Communist Party personnel data, how patron-client factions maintain cohesion, and Kim (2021) demonstrates empirically that the institutional cohesion and career trajectories of North Korean elites determine the risk of purge.

The insight that changes in elite composition are linked to policy lines is particularly evident in China studies. Lee (1991) and Joo (2009; 2017) show that the process by which China’s ruling elite shifted from ‘revolutionary cadres → technocrats → general bureaucrats’ was intertwined with the policy shift toward reform and opening. In other words, when the ‘type’ of elite changes, the ‘direction’ of the policy the regime prioritizes also changes.

2.2. North Korea as an “Empty” Space

The problem is that while this ‘elite agency theory’ has accumulated mainly around China, Russia, and Vietnam, where access to data is relatively easy, empirical analysis of North Korea remains very scarce. Han (2009; 2019; 2023) explains the advance and retreat of North Korea’s economic reforms through ‘organizational behavior’ and ‘bureaucratic politics’ models, pointing out the limits of ‘Supreme Leader determinism,’ and Ishiyama (2014) traces the turnover between hardliners and moderates through the network of on-the-spot guidance. Pyo & Heo (2019) capture the rise of the Party after 2016 in the party-government-military accompanying network during the Kim Jong Un era. However, most of these studies are limited to ‘describing’ the elite status at a specific point in time or fixing the network at a single point for analysis.

In summary, a distinct gap exists within the current literature. There has been almost no research that examines specific episodes of policy transition over time and compares the changes in elite composition before and after to trace the sequence between the two. As a first step to fill that gap, this study identifies four major shifts in strategic line during the Kim Jong Un era and traces the change in the composition of the ‘accompanying elite’ before and after each point in time, while also presenting a methodological agenda for identifying this.

3. Theoretical Framework: Power-Sharing and the Red-Expert Conflict

3.1. Power-Sharing

Svolik (2012) argues that the core conflict of authoritarian politics lies not in confrontation with the masses but in the problem of power-sharing between ‘the leader and the elite.’ According to his empirical findings, most dictators who lost power through unconstitutional means were ousted not by popular uprisings but by coups from within the elite. This implies that the most significant threats to an authoritarian leader originate from within their own ranks, rather than from external forces. Therefore, the leader cannot merely repress the elite but is compelled to share a certain degree of power in order to placate and co-opt them.

Applying this logic to policy-making, it becomes difficult to explain a major shift in line solely as the decision of a single leader. This is because a new policy brings benefits to some groups and losses to others. Therefore, before formalizing a policy, the leader may try to pre-emptively weaken the group that stands to lose and pre-emptively elevate the group needed to carry out the policy, in order to reduce the risk of backlash. Elites, too, can exercise agenda-setting power by first providing information and alternatives reflecting their own preferences, thereby shaping the Supreme Leader’s perception of the situation. As Meng et al. (2023) state, genuine power-sharing is not merely the distribution of benefits but includes ‘guarantees’ such as agenda-setting authority and coercive means. That both directions can be observed as ‘changes in the configuration prior to the announcement’ is the analytical starting point of this study.

3.2. Conflict Model and ‘Red and Expert’

Elite groups are never homogeneous. Groups that prioritize political loyalty and revolutionary zeal and groups that prioritize expertise and practical competence hold different policy preferences. This connects to the conflict model, which views authoritarian policy-making as the product of confrontation between groups, and more specifically to the policy tendency model, which views policy as ‘the outcome of competition among groups with differing policy preferences.’ Hamrin (1990) divided the Chinese elite into reformist and conservative camps and explained the direction of policy as the outcome of a contest between lines, while Moody (1984) conceptualized this as a confrontation between ‘Red’ and ‘Expert.’ The competition between ‘Red,’ which prioritizes revolutionary zeal and ideological orthodoxy, and ‘Expert,’ which prioritizes expertise, practical work, and economic performance, is said to encapsulate the conflict over policy line.

In the early, radical transformation phase of socialism, revolutionary cadres ("Red") set the political line and then staffed its execution. As the line shifted toward economic development and reform, however, demand surged for technocrats ("Expert") capable of managing a complex industrial society. Observing whether the ruling elite of a particular period leans toward ‘Red’ or ‘Expert’ becomes a key indicator for reading that regime’s policy priorities and power landscape.

Table 1. Concepts of ‘Red’ and ‘Expert’

3.3. Application to North Korea

This framework is valid for North Korea as well. On the surface, North Korean decision-making appears to be a vertical hierarchy, but in reality, organizations with differing interests—the Party, the military, and the Cabinet—are simultaneously involved in policy formation and implementation. The leader has no choice but to rely on the information reported by each organization, and that information inevitably carries the organization’s preferences.

North Korea’s economic reform phase in the 2000s is a good example. When the Cabinet pushed for greater use of markets and expanded autonomy through measures such as the July 1 Economic Management Improvement Measures (2002) and the official recognition of general markets (2003), the Party, concerned about the weakening of control, framed this as ‘excessive reform’ and moved to check it. In the end, reform retreated with the dissolution of the Cabinet’s Park Pong-ju task force (2005), the establishment of the Planning and Finance Department led by Pak Nam-gi (2005), and market controls through the June 18 statement (2008). This was less a policy failure than the result of inter-organizational conflict and the readjustment of power relations. In this configuration, the Cabinet, which valued pragmatism and the economy, can be seen as corresponding to ‘Expert,’ while the Party, which prioritized control and ideology, corresponds to ‘Red.’ In other words, the ‘Red-Expert’ policy conflict existed within North Korea as well.

This tension continues into the Kim Jong Un era. At the December 2022 Party Plenary Meeting, Kim Jong Un characterized the tendency of economic officials who valued external economic cooperation and the introduction of advanced technology as ‘defeatism and technological mysticism,’ rebuking them by saying that ‘outdated ideas remain latent.’ This episode demonstrates that, even amid prolonged sanctions, the tension between ‘Expert,’ which favors openness and technology, and ‘Red,’ which is wary of it, continues to operate.

4. Data and Methods

4.1. Why ‘Kim Jong Un On-the-Spot-Guidance Accompaniment Data’?

Existing research on the North Korean elite has generally read power changes through changes in official positions—that is, ‘who was appointed to which post.’ However, official personnel appointments are confirmed only long after the change has already progressed considerably. It is difficult to capture the subtle signs of substantive shifts in status.

For this reason, this study focuses on ‘data on accompaniment at Kim Jong Un’s public activities (on-the-spot guidance and events).’ The officials who accompany the Supreme Leader are not chosen at random. The policy needs of a given period determine who stays close to him, and accompaniment tends to appear earlier than official changes in position. In particular, on-the-spot guidance often involves long-duration and long-distance travel, providing ample opportunity for policy discussion between the Supreme Leader and accompanying elites. In other words, the attendance roster serves as a refined, real-time indicator of how the power mechanism operates.

Of course, the frequency of accompaniment does not perfectly measure influence itself. Ceremonial accompaniment may be mixed in, and there may be figures who wield real power despite accompanying less frequently. However, the goal of this study is not to measure the individual ranking of elites, but to observe, at the group level, the change in the overall proportion of accompaniment by the ‘Red’ or ‘Expert’ group before and after a transition. In this respect, the accompaniment data has sufficient analytical value. The data was constructed based on the Korea Institute for National Unification (KINU)’s ‘Database of Reports and Analysis on Kim Jong Un’s Public Activities,’ and missing or unclear items were cross-verified with Rodong Sinmun reports and the Ministry of Unification’s biographical data on figures (2012–2022).

4.2. ‘Red-Expert’ Coding

Another limitation of existing research is that it classifies elites mainly by ‘affiliated institution’—Party, government, or military. However, this classification method makes it difficult to discern preferences and roles in policy line when figures from different affiliations perform similar functions. Therefore, this study coded ‘Red’ and ‘Expert’ based not on affiliation but on the function and expertise the individual actually performed.

Each figure was coded as either ‘Red (1)’ or ‘Expert (0),’ and for figures who moved between the Party and the Cabinet, the judgment was based on their primary duty at the time of accompaniment, with actual work experience prioritized over educational background. To maintain coding consistency, individuals with specialized education were excluded from the 'Expert' category if they lacked relevant practical work experience.

4.3. Analytical Tools and Cautions in Identification

The primary analytical tool of this study is lead-lag regression. Using the four policy announcement points as the reference (0), an ‘elite-month’ panel of a total of 13 months—six months before and six months after—is constructed to estimate the trend in the proportion of ‘Red.’ If the Red-Expert proportion changes only after (lag) the announcement, this is judged to be closer to a ‘top-down’ pattern in which the Supreme Leader decided first and fit people to it afterward. Conversely, if the proportion had already changed before (lead) the announcement, this is judged to be closer to a ‘bottom-up’ pattern in which the elite configuration moved first and drove the decision.

However, two additional conditions would be needed for this design to definitively identify whether the change is ‘leading’ or ‘after the fact,’ and the current data does not yet sufficiently satisfy those conditions. First, to say that ‘the lead is higher than usual,’ there must be a pre-existing baseline point sufficiently distant from the announcement time. However, as it stands, looking only at a 13-month window around the policy announcement means that the entire 13 months constitutes the event window, so no baseline for comparison exists within the window. As a result, while the overall pattern can be identified, it is difficult to establish the ‘statistical significance’ of the lead with certainty. Second, because strategic line shifts are rare events—only four have occurred during the Kim Jong Un era—the statistical power of average-based inference is somewhat limited. Accordingly, it should be stated at the outset that the coefficients and significance levels in this presentation should be read not as definitive hypothesis testing but as exploratory patterns.

4.4. Object of Analysis: Four Shifts in National Strategic Line

The object of analysis is the four strategic line shifts formalized through Party Plenary Meetings and the Supreme People’s Assembly since Kim Jong Un took power: ① the parallel line of economic construction and nuclear force building (March 2013, 6th [Central Committee], 23rd Plenary Meeting), ② the line of concentrating all efforts on socialist economic construction (April 2018, 7th [Central Committee], 3rd Plenary Meeting), ③ the frontal breakthrough campaign (December 2019, 7th [Central Committee], 5th Plenary Meeting), and ④ the legislation of nuclear force (September 2022, 14th Supreme People’s Assembly, 7th Session). All four phases are periods in which the priority between the two pillars of regime survival—‘national defense (nuclear force)’ and ‘economic construction’—was readjusted, making them suitable for exploring Red-Expert dynamics.

5. Three Competing Hypotheses

This study sets up three competing hypotheses and explores, in light of the data, which interpretation is more valid. Hypothesis 1 reflects the traditional ‘Supreme Leader determinism’; Hypothesis 2 reflects ‘power-sharing/elite agency theory’; and Hypothesis 3 reflects ‘sequential power-sharing,’ in which the two are dynamically combined. The primary objective is not to definitively prove a single hypothesis, but rather to examine which mechanisms are most salient during specific phases.

Table 2. Three Competing Hypotheses

6. Analytical Results: Four Scenes

Figure 1 provides a comparative overview of the four phases. The first observation is that the pattern differs distinctly from phase to phase. That such markedly different results emerge from the same data and the same analytical method suggests that North Korea’s policy shifts are a dynamic political process that cannot be reduced to a single formula. It should be stated at the outset that the following four ‘scenes’ are not the results of definitive tests but exploratory sketches that describe the Red/Expert composition of each phase and interpret it together with context.

Figure 1. Comparison of Changes in the Proportion of Red-Expert Elite Accompaniment Before and After the Four National Strategic Line Shifts

6.1. Scene ① The Parallel Line (March 2013) — ‘Red’ Dominance Throughout the Period

Figure 2. Before and After the Announcement of the ‘Economy-Nuclear Parallel Policy’ 

(positive = Red dominance)

In the phase of the parallel line, the proportion of ‘Red’ elites is observed to be high throughout the period. Positive (+) values continue from six months before the announcement (lead6, coefficient 0.131) to just after the announcement (lag2, 0.178), before weakening from the fourth month (lag4) onward. The direction of this transition, which elevated nuclear force/national defense as one pillar, is consistent with the relative dominance of regime-defense (Red) elites.

6.2. Scene ② The Line of Concentrating All Efforts (April 2018) — The Most Robust Rise of ‘Expert’

Figure 3. Before and After the Announcement of the ‘Strategic Line of Concentrating 

All Efforts on Socialist Economic Construction’ (negative = Expert dominance)

This is the exact opposite picture. In this shift toward concentrating resources on economic construction, ‘Expert’ elites are prominent. Interestingly, this pattern is already clear in the period before the announcement—it appears that, before the ‘concentration on the economy’ declaration was formalized, economic and practically-oriented elites were already stationed in large numbers beside the Supreme Leader. Moreover, the rise of ‘Expert’ in this phase is the pattern most robustly confirmed statistically when the four phases are compared together. Recalling that the 2018 ‘concentration on the economy’ coincided with that year’s dramatic inter-Korean and North Korea–U.S. summit diplomacy, this rise of ‘Expert’ suggests the possibility that it was not merely a personnel matter but a harbinger of the line shift. This can be interpreted as a result supporting the bottom-up (Hypothesis 2) interpretation.

6.3. Scene ③ The Frontal Breakthrough Campaign (December 2019) — A Shift from ‘Expert’ to ‘Red’

Figure 4. Before and After the Announcement of the ‘Frontal Breakthrough Campaign’ 

(Expert → Red shift)

In the frontal breakthrough campaign, launched under the banner of ‘self-reliance’ amid a deadlock in sanctions, the two mechanisms appear to emerge sequentially. ‘Expert’ elites are dominant before and immediately after the announcement, but the weight shifts toward ‘Red’ from about three months after the announcement (lag3) onward. This can be read as a picture in which, during the phase when negotiation momentum still remained, practical-work and economic-type elites led the agenda, but as the deadlock solidified, the ruling coalition was reorganized around regime-cohesion-type elites.

6.4. Scene ④ Legislation of Nuclear Force (September 2022) — No Clear Change

Figure 5. Before and After the Announcement of the ‘Legislation of Nuclear Force’ 

(mostly non-significant)

In the fourth scene, no clear change is observed before or after the announcement. This measure, which codified nuclear doctrine into law, is closer to a ‘symbolic/declaratory’ policy that does not entail substantial reallocation of resources or reorganization of roles, and the ‘non-response’ of the elite configuration is consistent with that interpretation.

6.5. The Four Scenes in One Table

Table 3. Observed Patterns and Exploratory Interpretations of the Four Scenes

In sum, the four phases show differing Red/Expert composition profiles. The two points that can be stated most robustly are (a) the ‘structural persistence of Red dominance,’ in which ‘Expert’ fails to overtake ‘Red’ in most periods, and (b) the relative rise of ‘Expert,’ which stands out in the economically oriented phase (concentration of economic effort). Meanwhile, a more robust verification of the claim that ‘the elite configuration moved ahead of the announcement’ remains a task to be verified through additional design. What is clear is that the linear picture of ‘the Supreme Leader decides and the elite follows’ (Hypothesis 1) alone cannot explain this phase-by-phase heterogeneity.

7. Three Implications

1) Clues to Move Beyond Supreme Leader Determinism

The heterogeneous pattern across the four phases, and in particular the rise of ‘Expert’ in the economically oriented phase, suggests that even in North Korea—classified as the most closed one-man dictatorship—policy line and elite composition can be linked. This is not captured by a linear top-down picture alone, and it provides a clue for extending the ‘elite agency theory’ that comparative authoritarianism research has accumulated in China, Russia, and Vietnam to the case of North Korea, where data constraints are greatest.

2) The ‘Nature’ of a Policy May Determine Elite Dynamics

The elite configuration’s response may differ between policies that actually readjust resource allocation and priorities and policies that remain merely symbolic declarations. If this difference is robustly confirmed in follow-up analysis, whether the elite configuration responds could serve as an auxiliary yardstick for gauging whether a given announcement is ‘rhetorical’ or ‘substantive.’

3) The Announcement May Be a Result, Not a Starting Point — Potential as a Leading Indicator

If elite reorganization precedes policy announcement, then North Korea’s official announcements are closer to an ‘endpoint’ at which internal coordination has largely concluded. In that case, to read North Korea’s next moves, one should not wait for the announcement text but should preemptively track changes in the elite configuration around the Supreme Leader, and accompaniment data serves as compelling leading evidence for that early warning.

The message running through these three implications is as follows. North Korea is a ‘black box,’ but not total darkness. Subtle changes in who stands beside the Supreme Leader reveal clues about the direction and weight of policy that the announcement text does not tell us. If we can systematically collect and rigorously analyze these clues, we can read North Korea within the framework of comparative authoritarianism rather than isolating it as an ‘exceptional case,’ and at the same time obtain valid clues for our policy toward and intelligence analysis of North Korea.

Paradoxically, it is precisely now — with the 'two hostile states' framework firmly entrenched — that preparation matters most. Even amid severed exchange, the elite configuration within North Korea does not stop moving, and that movement, as seen in the accompaniment data, gradually leaks to the outside little by little. If we read these signals steadily, systematically, and with methodological rigor, then when the next inflection point arrives, we can stand not on the side that ‘reacts passively’ but on the side that ‘reads and designs first.’ This is precisely where the value of evidence-based North Korea studies lies.

8. Policy Implications and Recommendations

The results and approach of this study are not confined to academic contribution. They also offer direct implications for government agencies, intelligence analysis, the media, and other practitioners working on North Korea. Those implications converge on one point: we must read not ‘the North Korea we wish for’ but ‘North Korea as it is,’ on the basis of clear evidence. To this end, three recommendations are added.

1) From ‘The North Korea We Wish For’ to ‘North Korea As It Is’ — From Value-Laden Judgment to Evidence-Based Approach

Policy toward North Korea is often swayed by wishful thinking or political preference, portraying North Korea as ‘what we would like it to be.’ However, as this study shows, systematically analyzing observable signals such as accompaniment data allows for a more sober assessment of the actual power landscape and policy priorities within North Korea. For instance, the structural persistence of ‘Red’ dominance, which does not waver in any phase, tempers optimism of the kind that ‘reform and opening are imminent,’ while the relative rise of ‘Expert’ observed in the economically oriented phase provides a clue as to ‘under what conditions’ the pragmatic line gains strength. Understanding of North Korea ‘as an entity,’ grounded in such objective evidence, rather than a policy driven by normative expectation, should be the starting point for policy.

2) Cultivating Professional Personnel Who Can Interpret and Judge Objective Information

Regardless of the volume of high-quality data, it holds limited utility without the analytical capacity to interpret it. As this study shows, the same signal can lead to entirely different conclusions depending on the method of analysis and how its limitations are assessed. Therefore, government, intelligence agencies, and the media urgently need ‘methodologically trained’ professional personnel who can collect, accumulate, and interpret North Korea-related data and, on that basis, make careful policy judgments. These are personnel who combine area-studies knowledge with quantitative and qualitative methods, and who can critically assess the uncertainty of signals and the limits of identification. To this end, it is proposed that an interdisciplinary analyst training track be institutionalized, linking graduate education with in-service retraining.

3) Institutionalizing Data Collection and Accumulation for Predicting North Korean Behavior

The greatest constraint on this study was ultimately data limitation. That is, there were only four events and the observation window was short. This is not a problem for individual researchers but a problem of ‘research infrastructure.’ A national data infrastructure is needed to collect, standardize, and accumulate, over the long term and systematically, observable signals such as Kim Jong Un’s public activities, personnel appointments and purges, media discourse, and budget and trade data. Only data that is ‘continuously accumulated,’ rather than one-off analysis, can capture an inflection point ahead of time. A standard dataset and open codebook shared by KINU, the Ministry of Unification, intelligence agencies, and academia, along with a permanent system to update and manage it, is proposed.

These three recommendations are not separate but form a single cycle: gathering signals (data), cultivating people who can read them (personnel), and making judgments on the basis of evidence (evidence-based policy). Only when this cycle operates can we stand, at the next inflection point in inter-Korean relations, not on the side that ‘reacts passively’ but on the side that ‘reads and designs first.’ Even against a closed North Korea, evidence can be accumulated, and capacity can be cultivated.

9. Conclusion: Limitations and Next Steps

This study is at an exploratory stage and requires more robust data accumulation and analysis going forward. Several limitations to this study should be noted. First, regarding measurement constraints, the frequency of accompaniment does not fully measure influence (issues such as ceremonial accompaniment and real power-holders who accompany less often). However, since the interest lies not in individual ranking but in the trend of group-level proportions, this problem is partially mitigated. Second, there is subjectivity in coding. The Red-Expert distinction involves a certain degree of judgment in interpreting careers. We plan to transparently disclose the coding rules, supplement them with cross-verification, and report double-coding reliability (κ) in the future. Third, there is a limitation of identification. The ±6-month window covers the entire event period, so there is no suitable prior baseline point, and because there are only four national strategic line shifts in total, the statistical identification and generalization of dynamic effects are inherently limited. Therefore, the ‘lead’-related interpretation in this presentation is not a definitive conclusion but an exploratory interpretation.

The next stage of research should proceed in a way that overcomes these limitations. First, the time window should be expanded to ±12 months before and after the announcement to secure a local prior baseline, and the four phases should be combined to estimate the average change relative to a common baseline with valid standard errors. This would separate whether the ‘movement of the lead’ is a leading effect or a phase-level difference. Next, one should move beyond group-level proportions to individual-level entry-exit dynamics (survival analysis) to directly trace ‘who rises first.’ Also, acknowledging the inherent constraint that strategic line shifts are rare events, a mixed design combining quantitative techniques with process tracing (Rodong Sinmun, Plenary Meeting documents, tracking of individuals) should identify the causal sequencing mechanism within the case. Finally, it is also necessary to combine accompaniment data with other signals such as personnel appointments/purges and media discourse, and, in the long term, to extend it to time-series causality with text-based policy-orientation indices, broadening the scope of ‘evidence-based North Korea studies.’

Extreme opacity is an inherent constraint in North Korean studies. Yet if we collect, systematically and with methodological rigor, the signals that leak through its cracks, we can read the politics that lies beyond the Supreme Leader's instructions. The winter of severance between South and North Korea appears long, but it will not last forever. And the side that reads the inflection point of spring first will be the side that has gathered its signals in advance, systematically.  ■ 

References

Bueno de Mesquita, Bruce, Alastair Smith, Randolph M. Siverson, and James D. Morrow. 2003. The Logic of Political Survival. Cambridge, MA: MIT Press.

Geddes, Barbara, Joseph Wright, and Erica Frantz. 2018. How Dictatorships Work. Cambridge: Cambridge University Press.

Hamrin, Carol Lee. 1990. China and the Challenge of the Future. Boulder, CO: Westview Press.

Han, Ki-beom. 2009. “The Pursuit and Frustration of North Korean Economic Reform: Applying the Organizational Behavior and Bureaucratic Politics Models.” PhD diss., Graduate School of Korea University.

_____. 2019. The Pursuit and Frustration of North Korean Economic Reform: The Organizational Behavior and Bureaucratic Politics Models. Seoul: Hanul Academy.

_____. 2023. North Korean Bureaucratic Politics and Economic Policy in the Kim Jong Un Era. Seoul: Sunin.

Ishiyama, John. 2014. “Assessing the Leadership Transition in North Korea: Using Network Analysis of Field Inspections, 1997–2012.” Communist and Post-Communist Studies 47(2): 215–226.

Joo, Jang-hwan. 2009. “A Study on the Recruitment of the Chinese Political Elite and Changes in Policy Line: Focusing on the 16th Central Committee Members.” Journal of Sino-Soviet Affairs 33(2): 145–182.

_____. 2017. “Generational Change of the Chinese Political Elite and Policy Change.” Journal of Chinese Studies 81: 213–242.

Kim, Taekbin. 2021. “Who Is Purged? Determinants of Elite Purges in North Korea.” Communist and Post-Communist Studies 54(3): 73–96.

Korea Institute for National Unification (KINU). 2024. “Database of Reports and Analysis on Kim Jong Un’s Public Activities.” Seoul: Korea Institute for National Unification.

Lee, Hong Yung. 1991. From Revolutionary Cadres to Party Technocrats in Socialist China. Berkeley: University of California Press.

Meng, Anne, Jack Paine, and Robert Powell. 2023. “Authoritarian Power Sharing.” Annual Review of Political Science 26(1): 183–204.

Ministry of Unification. 2022. Biographical Information on Major North Korean Figures (2012–2022). Seoul: Ministry of Unification.

Moody, Peter R. 1984. “Red vs. Expert.” Asian Survey 24(8): 825–841.

Newson, Adlai, and Francesco Trebbi. 2018. “Authoritarian Elites.” Canadian Journal of Economics 51(4): 1088–1117.

Pyo, Yun-shin, and Heo, Jae-young. 2019. “An Analysis of the Elite Network Accompanying On-the-Spot Guidance during the Kim Jong Un Era.” Contemporary North Korean Studies 22(2): 7–45.

Shih, Victor, and Jonghyuk Lee. 2023. “Factional Survival in the Chinese Communist Party.” Comparative Political Studies 56(3): 331–362.

Svolik, Milan W. 2012. The Politics of Authoritarian Rule. Cambridge: Cambridge University Press.

■ Taekbin KIM is an Assistant Professor at the Korea National Defense University.

■ Translated and edited by Sangjun LEE, EAI Research Associate
    For inquiries: 02 2277 1683 (ext. 211) | [email protected]