Date of Award

Fall 1-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Political Science

First Advisor

Mattingly, Daniel

Abstract

How does political communication function when the channels between elites, media, and the public become fragmented and unpredictable? This dissertation investigates how influence is exercised in a shifting media environment, where traditional broadcast-era assumptions about information flow and audience exposure no longer hold. Rather than assuming stable pathways from elite messaging to public response, the projects examine how political actors, media institutions, and citizens operate under new constraints and opportunities for choice. It asks whether elites can still shape media tone in ways that affect public support, whether media outlets retain agenda-setting power in a competitive landscape, and how citizens respond to political content under conditions of selective exposure and emotional disaffection. Theoretically, the project rethinks political communication as a dynamic process of influence shaped by fragmentation. It argues that the decline of centralized media weakens elites’ ability to shape public opinion through agenda-setting and instead elevates the importance of tone management through media outreach, or soft capture. At the same time, agenda-setting authority itself becomes segmented across media types and topics, with alternative sources gaining influence in areas where public demand is high and institutional supply is weak, such as criticism of central government. Finally, the project assesses the theory of `fit': the alignment between individual disposition and media content—as a key determinant of audience response in environments of media choice. The project tests these theoretical claims through three empirical studies. The first paper uses content analysis of over 5 million Japanese-language news articles (2004–2024), linked to cabinet approval and prime minister schedule data to assess how elite engagement with media shapes tone and public opinion. The second applies large-scale machine learning analysis to link news articles and tweets, tracing how narratives circulate across platforms and identifying when alternative sources gain narrative dominance. The third presents a survey experiment using the Preference-Incorporating Choice Assignment II framework to assess how fit and disaffection shape affective and attitudinal reactions. Findings show that favorable tone in legacy media is associated with increased cabinet support and that elite outreach moderates this tone, though this influence declines over time. Legacy media remain central to political discourse overall, but have lost dominance in critical or anti-government narratives. Citizens respond more strongly to non-preferred media content than to coverage they would have chosen, especially when disaffection is high, suggesting that emotional orientation can outweigh partisanship in shaping reactions to media. Together, these results revise existing models of political communication by highlighting the conditional, selective, and context-dependent nature of influence in fragmented systems. They show that political communication still matters, but that its effects depend on who is speaking, who is listening, and how well their signals align. As such, the project contributes to a more nuanced and adaptive theory of influence under hybrid media conditions.

Download

Share

COinS