How Diplomacy Reshapes Online Discourse:Asymmetric Persistence in Online Framing of North Korea

TL;DR

The paper investigates how high-stakes diplomatic summits reframe online discourse toward North Korea by separating content-level framing from structural network changes. Using a Difference-in-Differences design with multiple global controls, an LLM-based Codebook framing classifier, and GraphRAG-derived discourse networks on Reddit data (2017–2019), it finds that framing shifts toward diplomacy are robust and partially persistent after Hanoi’s failure, while sentiment reverts fully. The analysis reveals multi-level discourse reorganization: edges, communities, and overall network cohesion shift toward diplomacy-oriented structures that endure beyond diplomatic setbacks. Audience responses corroborate these effects, indicating propagation into comments and networks. The study demonstrates that diplomacy can durably alter interpretive frames in online publics, with implications for diplomacy and strategic communication.

Abstract

Public opinion toward foreign adversaries shapes and constrains diplomatic options. Prior research has largely relied on sentiment analysis and survey based measures, providing limited insight into how sustained narrative changes (beyond transient emotional reactions) might follow diplomatic engagement. This study examines the extent to which high stakes diplomatic summits shape how adversaries are framed in online discourse. We analyze U.S.-North Korea summit diplomacy (2018-2019) using a Difference-in-Difference(DiD) design on Reddit discussions. Using multiple control groups (China, Iran, Russia) to adjust for concurrent geopolitical shocks, we integrate a validated Codebook LLM framework for framing classification with graph based discourse network analysis that examines both edge level relationships and community level narrative structures. Our results reveal short term asymmetric persistence in framing responses to diplomacy. While both post level and comment level sentiment proved transient (improving during the Singapore Summit but fully reverting after the Hanoi failure),framing exhibited significant stability: the shift from threat oriented to diplomacy oriented framing was only partially reversed. Structurally, the proportion of threat oriented edges decreased substantially (48% -> 28%) while diplomacy oriented structures expanded, and these shifts resisted complete reversion after diplomatic failure. These findings suggest that diplomatic success can leave a short-term but lasting imprint on how adversaries are framed in online discourse, even when subsequent negotiations fail.

Figures (5)

• Figure 1: Research Timeline and Key Events. Dashed brackets indicate analysis periods: P1 (Pre-Announcement), P2 (Singapore-Hanoi), P3 (Post-Hanoi). Transition periods are excluded from analysis to avoid anticipation effects.
• Figure 2: Sentiment Difference-in-Differences Visualization. Pre-treatment trends (P1) are visually parallel across groups, supporting the validity of the DiD design. The bold black dashed line marks the intervention point (summit date).
• Figure 3: Framing Difference-in-Differences Visualization. Pre-treatment trends are visually parallel for China and Iran, validating their use as control groups. Russia is excluded from (A) due to parallel trends violation (p=0.01) but included in (B) where parallel trends are satisfied. The bold black dashed line marks the intervention point (summit date).
• Figure 4: Monthly Sentiment Score Trends: NK vs. Control Groups. Vertical dashed lines mark key diplomatic events: the green line indicates the Singapore Summit (June 2018), and the red line indicates the Hanoi Summit (February 2019).
• Figure 5: Monthly Framing Score Trends: NK vs. Control Groups. Vertical dashed lines mark key diplomatic events: the green line indicates the Singapore Summit (June 2018), and the red line indicates the Hanoi Summit (February 2019).