From Deliberation to Affect: The Transformation of Public Policy Discourse in Indonesia’s Hybrid Media System (2020–2024)

Abstract

The displacement of Indonesia’s capital city (IKN) provides a critical case for examining how long-term public policy discourse is processed within a fragmented, algorithm-driven hybrid media system. Traditional agenda-setting theory, which attributes primary agenda power to mass media, cannot adequately explain this transformation. We therefore investigate the cross-platform dynamics of IKN discourse across online news media, Twitter/X, and TikTok over five years (2020–2024), hypothesizing a shift from text-based deliberation to visual-affective communication. Our sequential mixed-methods design combines longitudinal big data analysis—covering 368,291 news articles, 2.3 million Twitter/X posts, and 164,025 TikTok videos—with Granger causality tests within VAR models to trace temporal lead-lag relationships across five key policy events. The results reveal a punctuated, non-linear pattern of public attention. Twitter/X discourse peaked in 2022 around legislative debates, then declined by approximately 50% in 2023. Conversely, TikTok engagement grew exponentially, reaching 29.6 million interactions by 2024. Daily Spearman correlations show strong intersocial synchronization, with the highest pairwise correlation between Twitter/X and TikTok (ρ = 0.765, p < 0.001). Granger causality tests demonstrate event-contingent agenda-setting: ceremonial events (e.g., the August 17th celebration at IKN) exhibit bidirectional causality between online media and TikTok, while deliberative events show top-down influence from online media to both platforms. Notably, we observe a reverse intermedia agenda-setting phenomenon during the August 2024 celebration, where high engagement on TikTok preceded and drove increased coverage in national online media. These findings support our central claim: agenda-setting for national strategic projects has fundamentally transformed from a deliberative model to an affective political model mediated by platform algorithms. We introduce the concept of the Visual Resonance Engine to describe how TikTok compels emotional visual content to preempt and reshape the editorial logic of online journalism. This research contributes to understanding how algorithmic logic intervenes in news priorities, highlighting the shift from rational argumentation to performative visual resonance as the dominant mode of public policy communication.