Risk Perception and Media Polarization in International Environmental Disputes: The Fukushima Wastewater Issue in South Korea

Abstract

This study examines how media outlets perceive and frame environmental issues, focusing specifically on South Korean news reporting on the Fukushima wastewater release. Using computational text analysis methods and formal statistical testing, this research investigates ideological polarization in media framing and its implications for risk perception. Statistical analyses provide significant evidence that progressive media emphasized environmental risks, public health concerns, and governmental accountability, whereas conservative media highlighted political actors, institutional governance, and diplomatic considerations. Media polarization peaked before policy implementation but moderated afterward. After implementation, moderate newspapers exhibited ideological convergence toward the center, accounting for a substantial portion of the observed positioning shifts, while strongly partisan outlets maintained or slightly intensified their prior positions. These results suggest that ideological polarization is most prominent in the early stages of contentious environmental policies but moderates as empirical scrutiny replaces speculation, with varying effects across ideological groups. This study provides insights into the temporal dynamics of media polarization and its role in shaping risk perception during policy implementation and regulatory oversight.

1. Introduction

The 2011 meltdown at Japan’s Fukushima Daiichi Nuclear Power Plant, following a massive earthquake and tsunami, became a critical moment in global environmental and political discourse. Over a decade later, Japan’s decision to release Advanced Liquid Processing System (ALPS)-treated radioactive water into the Pacific Ocean in August 2023 has reignited international debate, particularly in neighboring South Korea. This decision, supported by the International Atomic Energy Agency (IAEA) as meeting global safety standards, has sparked polarized responses in South Korea, reflecting historical grievances and ideological divides. In South Korea, media coverage of the Fukushima water release has been particularly contentious, with left-leaning outlets condemning the move as reckless and ecologically dangerous, while right-leaning media emphasize scientific assurances and criticize alarmist narratives as politically motivated. These divergent perspectives highlight the role of media bias in shaping public understanding and discourse on environmental issues.

While previous studies have explored media framing and bias in environmental reporting, this paper uniquely examines the ideological polarization of multiple media outlets by not only measuring their ideological positions but also tracking shifts in these positions over time. The primary objective of this study is to investigate how ideological polarization among South Korean media outlets influences their framing of environmental issues, specifically through the case of the Fukushima wastewater release. Specifically, this research addresses two central research questions: How does media framing of the Fukushima wastewater release differ between politically left-leaning and right-leaning outlets in South Korea? Under what conditions does media polarization on environmental issues become less severe? Based on theoretical frameworks such as Risk Perception Theory and the Issue–Attention Cycle, this study hypothesizes that media polarization will be most pronounced before policy implementation due to heightened uncertainty and speculative discourse but will subsequently moderate as empirical scrutiny replaces speculation. In addition, progressive media outlets are expected to predominantly frame the Fukushima wastewater release in terms of environmental and governance risks, whereas conservative outlets are more likely to emphasize scientific validation and international regulatory approval.

By testing these hypotheses, this study contributes to the understanding of media polarization in environmental communication. In this study, media polarization is defined as the variation in perspectives presented by different media outlets on specific issues. A critical issue here is the media’s gatekeeping process, where certain issues are selectively highlighted or excluded, and interpretations of events are shaped before being communicated to the public [1,2,3]. In addition, this study provides practical insights for policymakers, environmental advocates, and media practitioners seeking to promote balanced and accurate reporting on global environmental challenges.

The paper proceeds as follows: It first reviews the literature on media polarization and environmental journalism. It then provides the background and arguments surrounding the Fukushima wastewater release, presenting the theoretical framework. After describing the data and methodology, the study presents the empirical results. The paper concludes with a discussion of the implications for media polarization, environmental risk perception, and public discourse on transboundary environmental issues.

2. Existing Literature

Media coverage of political and environmental issues has long been central in communication studies, particularly focusing on the phenomenon of media polarization—systematic differences in news framing shaped by ideological orientations. Polarization becomes especially evident in debates surrounding scientific and environmental controversies [4]. Theoretical frameworks like agenda-setting illustrate that media influence public discourse by determining issue prominence [5], while framing theory emphasizes how narratives constructed by journalists influence audience interpretations of policy issues [1].

The substantial body of literature demonstrates the significant impact of ideological biases on media framing. Nisbet [6] highlights that media frames substantially shape public understanding and policy support for environmental issues by altering audience perceptions. Further reinforcing this view, D’Alessio and Allen’s [7] comprehensive meta-analysis reveals pervasive ideological biases in news coverage, with systematic framing variations reflecting political orientations.

Several structural and ideological factors underpin media polarization. Primarily, media narratives align with the ideological orientations of news outlets. However, factors such as national security concerns, economic interests, and diplomatic pressures also affect how these biases manifest, resulting in varied framing strategies across different contexts [8]. Partisanship significantly contributes to these patterns, with conservative and progressive media emphasizing distinct aspects of identical issues [9]. Moreover, selective exposure, whereby audiences prefer news aligning with their existing beliefs, reinforces these ideological divides [10]. Economic imperatives also influence editorial decisions as media organizations compete for audience loyalty and financial sustainability [11].

Environmental journalism offers a clear illustration of these dynamics, where scientific issues are often portrayed through ideological lenses. Boykoff and Boykoff [12] introduced the “balance as bias” concept, demonstrating how journalistic norms of neutrality might distort public understanding by disproportionately highlighting minority scientific viewpoints. Studies of climate change reporting show that conservative media frequently emphasize economic costs and uncertainties, whereas progressive media underscore scientific consensus and advocate regulatory action [13]. Such framing differences profoundly influence public perceptions of environmental risks and attitudes toward policy interventions [14].

Studies centered on South Korea specifically reinforce these general findings. Lee et al. [15] investigated how media platforms and political orientations affect climate change beliefs and pro-environmental behaviors among South Korean audiences, underscoring the strong influence of ideologically driven media exposure on environmental attitudes. Yun et al. [16] explored climate policy networks in South Korea, identifying distinct patterns of alliances and conflicts among media, government, and civic groups shaped by ideological positions. Their findings illustrate the intricate connections between media framing, political ideologies, and environmental governance.

The relationship between media framing and public risk perception is particularly salient in debates over nuclear energy. Perko [17] argues that media narratives critically shape public evaluations of nuclear risks. Nuclear energy and related disasters consistently rank among the most ideologically polarized environmental issues. Gamson and Modigliani [18] described competing cultural frames surrounding nuclear energy, with narratives ranging from symbols of technological advancement to existential environmental threats. Culley et al. [19] analyzed media portrayals of nuclear energy’s economic benefits and risks, highlighting media’s pivotal role in public perception. Vossen [20] examined Dutch media framing of nuclear energy within climate change discourse, uncovering traditional danger-focused narratives alongside emerging frames such as ecomodernism.

Emerging computational analyses further clarify the dynamics of media framing in nuclear disasters. Lansdall-Welfare et al. [21] examined news coverage following the Fukushima disaster, revealing shifts in sentiment and narrative as the issue transitioned from immediate crisis to sustained policy debate. Similarly, Zarrabeitia-Bilbao et al. [22] analyzed social media discussions on nuclear energy, finding that public sentiment evolved from emotional reactions toward more policy-centered debates over time.

Although media polarization remains deeply entrenched, certain conditions can moderate ideological divides. Feldman et al. [23] suggest ideological gaps in coverage narrow when policy issues are perceived as immediate and severe threats. Clear technical communication and robust regulatory oversight also mitigate partisan differences in risk perception [6]. Schudson [24] contends that journalistic norms combined with scientific consensus contribute to more balanced reporting across ideological divides.

Despite substantial scholarship on media polarization and environmental coverage, significant gaps persist. Few studies have systematically examined ideological influences on media framing of transboundary environmental issues, particularly those involving interstate interests and geopolitical dynamics. Moreover, there is limited understanding of how media polarization evolves before and after the implementation of environmental policies, especially concerning nuclear risk governance. The Fukushima wastewater release offers an opportunity to investigate these gaps, examining how media narratives adapt to transboundary environmental controversies amid regulatory debates and geopolitical tensions. Addressing these gaps promises deeper insights into the evolution of media framing in response to contentious environmental policies.

3. Fukushima Wastewater Release and Media Polarization in South Korea

The Fukushima Daiichi nuclear disaster, triggered by the Great East Japan Earthquake on 11 March 2011, remains one of the most consequential nuclear accidents in modern history. The earthquake and tsunami led to reactor meltdowns, widespread radioactive contamination, and long-term environmental and political challenges. Over the years, the Japanese government and the Tokyo Electric Power Company (TEPCO), based in Tokyo, Japan, sought long-term solutions for containing the contaminated water. In April 2021, Japan formally announced its decision to release Advanced Liquid Processing System (ALPS)-treated wastewater into the Pacific Ocean, citing limited storage capacity and adherence to international regulatory safety standards [25]. Subsequently, the first discharge began in August 2023, with subsequent releases continuing in scheduled phases. As summarized in

Table 1. Chronology of Fukushima wastewater release (2011–2023).

Date	Event
11 March 2011	The Great East Japan Earthquake triggers the Fukushima Daiichi Nuclear Power Plant accident, leading to radioactive material leakage.
13 April 2021	The Japanese government officially announces its decision to release treated wastewater into the ocean.
24 August 2023	The first discharge of treated wastewater begins under IAEA monitoring.
11 September 2023	The first discharge phase is completed, releasing approximately 7788 tons of treated wastewater.
5 October 2023	The second discharge phase begins, targeting around 7810 tons of wastewater.
23 October 2023	The second discharge phase is completed as planned.
2 November 2023	The third discharge phase starts, with an expected release of approximately 7753 tons.
20 November 2023	The third discharge phase concludes, completing the scheduled release.
31 December 2023	By the end of 2023, a total of approximately 23,400 tons of treated wastewater had been released.

Source: International Atomic Energy Agency (IAEA) (2023) [25].

, these events unfolded over a period of more than a decade, reflecting the progression from the 2011 Fukushima disaster to Japan’s decision to release treated wastewater and the subsequent implementation of this policy.

Figure 1 visualizes fluctuations in media attention surrounding the Fukushima wastewater release. Media engagement peaked in April 2021, when Japan announced its decision to release ALPS-treated water, and surged again in August 2023 with the first discharge. The decline in coverage following these peaks suggests a transition from speculative discourse to regulatory monitoring. Collectively, this pattern indicates that media attention was event-driven, responding to key policy announcements.

Japan’s decision to release ALPS-treated wastewater sparked intense debate, with media outlets constructing divergent narratives that reflected ideological divisions over nuclear energy, environmental governance, and trust in regulatory oversight. Notably, the Fukushima wastewater release serves as a case study of how media polarization is most pronounced prior to policy implementation and tends to moderate afterward. Before the release, progressive and conservative media framed the issue in starkly different ways, reflecting their respective ideological commitments to nuclear policy, confidence in regulatory oversight, and environmental risk perception. During the lead-up to implementation, uncertainty dominated media discourse, thereby creating fertile ground for ideologically polarized coverage. However, as the policy was enacted, media narratives shifted away from ideological contestation toward a more procedural, regulatory-focused discourse.

While progressive outlets emphasized environmental and governance failures, conservative outlets framed the release as a scientifically justified and internationally validated policy decision. In particular, progressive media framed the wastewater release as an environmental and governance failure by emphasizing potential long-term contamination risks, public opposition, and a lack of trust in both TEPCO and the Japanese government. In doing so, they amplified concerns from fishing communities, environmental NGOs, and neighboring governments by constructing a precautionary narrative—a framing strategy in which the absence of definitive scientific consensus justifies opposition to policies with uncertain risks [26]. This approach is consistent with past media responses to nuclear and environmental controversies, where uncertainty has been leveraged to advocate for stricter regulatory oversight and precautionary policymaking [13].

In contrast, conservative media framed the policy as a scientifically validated and internationally approved step toward nuclear waste management, relying on assessments from authoritative bodies such as the IAEA, government reassurances, and the necessity of efficiently managing post-disaster recovery. They sought to counteract public skepticism by reinforcing trust in international regulatory institutions, thereby portraying opposition as driven primarily by political motives rather than scientific evidence. By characterizing concerns from environmental activists and neighboring governments as exaggerated or unfounded, conservative outlets positioned the wastewater release as a pragmatic solution instead of an unresolved environmental risk. This framing aligns with Gamson and Modigliani’s [18] “progress” frame, which portrays nuclear energy as a symbol of technological advancement and economic development by emphasizing its benefits while downplaying potential risks.

The intensity of polarization before implementation is consistent with the Social Amplification of Risk Framework, which explains how uncertainty and political contestation heighten public concern, especially when media outlets amplify competing risk narratives [27]. Media engagement during this stage was highly emotional, focusing on worst-case scenarios, public outrage, and governmental accountability, thereby reinforcing audience segmentation based on ideological predispositions. Consequently, the pre-release period experienced strong partisan contestation, with risk perception shaped more by political and media narratives than by scientific consensus.

As the wastewater release proceeded, media framing gradually shifted away from polarized risk narratives toward a more neutral, regulatory-focused discourse. This shift aligns with normalization processes observed in environmental governance [28]. Studies on nuclear waste disposal, industrial pollution, and climate change policies have shown that intense public and media opposition tends to subside once a controversial environmental policy is enacted—unless immediate harm is observed. A similar pattern emerged in the Fukushima case, where pre-release coverage emphasized uncertainty, potential environmental degradation, and transnational opposition, while post-release discourse increasingly incorporated scientific monitoring data and ongoing regulatory oversight.

This transformation also reflects the Issue–Attention Cycle, which predicts that media engagement with controversial policies follows a trajectory in which initial alarm and polarization gradually give way to declining public interest and more procedural discussions [29]. Prior to the wastewater release, speculation about its environmental consequences drove high-intensity coverage, thereby sustaining ideological divisions. However, as implementation proceeded without an immediate ecological catastrophe, the urgency of the issue diminished, and media discourse shifted toward monitoring reports and technical assessments rather than existential debates. As overall media coverage declined, the opportunity for sustained ideological contestation diminished, resulting in a transition toward regulatory discussions rather than partisan conflict.

Risk Perception Theory further explains why the absence of visible harm led to a reduction in media contestation [26]. Prior to implementation, hypothetical risks were more salient, framing the wastewater release as a high-stakes environmental decision. However, once the policy was enacted and empirical observations did not align with worst-case projections, public concern and media intensity decreased accordingly. This shift mirrors patterns observed in previous nuclear-related policies, where media engagement tends to peak in response to uncertainty and subsequently moderates as outcomes become more predictable. To empirically evaluate this transformation in media framing, the next section applies computational methods to analyze how media outlets positioned themselves before and after the wastewater release.

These theoretical perspectives highlight how media polarization is shaped by ideological commitments and how it evolves in response to policy implementation. Building on this framework, this study formally tests the following hypotheses:

Hypothesis 1 (H1).

Media polarization regarding the Fukushima wastewater release will peak before policy implementation and moderate afterward.

Hypothesis 2 (H2).

Progressive media outlets will predominantly frame the Fukushima wastewater release in terms of environmental and governance risks, whereas conservative outlets will emphasize scientific validation and international regulatory approval.

To empirically evaluate these hypotheses regarding the dynamics and ideological characteristics of media polarization, the next section applies computational methods to analyze how media outlets positioned themselves before and after the wastewater release.

4. Data and Method

This study analyzes media coverage of the Fukushima wastewater release by collecting a total of 7738 news articles published between 1 April 2021 and 31 December 2023 from ten major South Korean daily newspapers: Chosun Ilbo, Munhwa Ilbo, Segye Ilbo, DongA Ilbo, JoonAng Ilbo, Kukmin Ilbo, Seoul Shinmun, Hankook Ilbo, Kyunghyang Shinmun, and Hankyoreh. These outlets cover diverse topics—including politics, economics, and social and environmental issues—and represent a wide ideological spectrum from conservative (e.g., Chosun Ilbo, Munhwa Ilbo) to progressive (e.g., Kyunghyang Shinmun, Hankyoreh). All selected newspapers maintain established print editions and actively updated online platforms, ensuring comprehensive coverage in both traditional and digital formats. The research period includes the pre-release phase (April 2021–July 2023), characterized by anticipatory media coverage and public debate, and the post-release phase (August–December 2023), reflecting media reactions after policy implementation. This temporal distinction enables an empirical evaluation of how media narratives and ideological framings evolved in response to the wastewater release.

To ensure comprehensive coverage, all articles referencing the Fukushima wastewater issue were systematically web-scraped from digital newspaper archives using automated scripts written in Python (version 3.12.1). Articles were retrieved through keyword-based queries, including terms related to “Fukushima”, “wastewater”, “contaminated water”, “treated water”, “discharge”, “radiation”, “tritium”, “International Atomic Energy Agency (IAEA)”, and “Tokyo Electric Power Company (TEPCO)”. To enhance data reliability, the extraction process accounted for terminological variations across different media outlets. After retrieval, the dataset underwent text preprocessing, including deduplication, stopword removal, tokenization, and stemming, to improve analytical consistency.

Following preprocessing, the dataset was analyzed in R using computational text analysis techniques and statistical hypothesis testing. The analysis employed Wordfish modeling and word cloud visualization as primary computational methods. Word cloud analysis provides a visual representation of lexical salience, highlighting dominant themes and linguistic emphasis in progressive and conservative media coverage. In contrast, the Wordfish model quantitatively estimates ideological positioning, mapping media outlets along a latent ideological spectrum based on word frequency distributions.

The Wordfish model is based on a Poisson regression framework, where word occurrences in a document are modeled as a function of an underlying ideological dimension. To maintain comparability across different time periods, the estimated ideological scores were standardized, allowing observed shifts in ideological positioning to reflect substantive changes in media narratives rather than differences in scale. The equation for the Wordfish model is as follows:

$$\ln \left(y_{ij}\right)={\alpha }_i+{\psi }_j+{\beta }_i\times {\theta }_i,$$

where

• $y_{ij}$: the count of word j in document i.
• ${\alpha }_i$: a document-specific intercept term capturing the overall verbosity or length of document i.
• ${\psi }_j$: a word-specific intercept term representing the baseline frequency of word j across all documents.
• ${\beta }_i$: the discrimination parameter for word j, indicating how strongly the frequency of this word correlates with the latent ideological position ${\theta }_i$.
• ${\theta }_i$: the latent ideological position of document i on the estimated ideological scale.

To extend the analytical approach beyond descriptive techniques, formal statistical tests were implemented for both hypotheses. This methodological triangulation enhances the validity of findings by combining descriptive visualization, computational modeling, and inferential statistics.

For testing Hypothesis 1, a two-stage analysis of ideological positioning was conducted. First, changes in each newspaper’s ideological distance from the center were calculated using standardized Wordfish estimates before and after the release. Second, an Analysis of Variance (ANOVA) compared these changes across four ideological groups (strongly conservative, moderately conservative, moderately progressive, and strongly progressive). Post hoc pairwise comparisons with Bonferroni correction were performed to identify specific between-group differences, while eta-squared (η²) was calculated to assess effect size.

For testing Hypothesis 2, an environmental–scientific framing ratio metric was developed to quantify the relative emphasis on different framing approaches. Two distinct frame dictionaries were created as follows: (1) environmental and governance risk frame and (2) scientific validation and international approval frame. The environmental–scientific ratio was calculated for each newspaper as the proportion of environmental frame words relative to total frame words. Statistical testing employed bootstrap confidence interval analysis and effect size calculation (Cohen’s d) to identify ideological differences in framing patterns. By integrating descriptive visualization with inferential statistical modeling and hypothesis testing, this approach provides a comprehensive analysis of framing differences and ideological bias in media reporting on the Fukushima wastewater release.

5. Empirical Results

To empirically test Hypothesis 1 and Hypothesis 2, this study conducted several computational analyses and statistical tests examining how South Korean media covered the Fukushima wastewater release.

5.1. Media Polarization Dynamics Before and After the Wastewater Release

The analysis of ideological positioning reveals distinct patterns of media polarization and its evolution following the Fukushima wastewater release. Figure 2a presents the ideological positioning of ten major South Korean newspapers from April 2021 to December 2023, based on a Wordfish analysis of their coverage. The results indicate a clear ideological distribution, with progressive outlets (Hankyoreh, Kyunghyang Shinmun) positioned at the leftmost end, while conservative outlets (Chosun Ilbo, Munhwa Ilbo) appear at the rightmost positions. This distribution suggests that media coverage of the Fukushima wastewater release largely aligned with pre-existing ideological tendencies rather than reflecting a uniform narrative across outlets.

To classify media outlets with moderate ideological positioning, this study employs the interquartile range (IQR) approach, which defines the moderate range as the middle 50% of the ideological distribution. This method is particularly suitable for analyzing media ideological positions as it provides a robust, distribution-free approach that is less sensitive to extreme values than traditional standard deviation-based methods while offering natural boundaries for moderate positions [30]. As Figure 2b shows, the moderate range, defined by IQR, spans from −0.75 to 0.76. Based on this classification, the remaining newspapers fall within this range, indicating that their reporting was less ideologically polarized than the strongly progressive (Hankyoreh, Kyunghyang Shinmun) and strongly conservative (Chosun Ilbo, Munhwa Ilbo) outlets. These findings establish a baseline of media polarization prior to examining shifts following the wastewater release.

Figure 3a,b illustrate changes in ideological positioning among major South Korean newspapers before and after the Fukushima wastewater release. In Figure 3a, individual newspaper trajectories indicate that progressive (Hankyoreh, Kyunghyang Shinmun) and conservative (Chosun Ilbo, Munhwa Ilbo) outlets maintained relatively stable ideological positions, while moderate newspapers (Hankook Ilbo, Seoul Shinmun, Kukmin Ilbo, JoongAng Ilbo, Segye Ilbo, and Dong-A Ilbo) exhibited a shift toward the ideological center. This suggests a reduction in ideological polarization among moderate media outlets in the post-release period. Figure 3b further validates this trend at the group level, showing that while strongly ideological newspapers remained stable, both moderate progressive and moderate conservative outlets moved closer to the center. This group-based analysis highlights the asymmetry in ideological shifts: moderate newspapers adjusted their framing post-release, whereas ideologically extreme outlets largely retained their pre-existing positions.

To formally test these observed convergence patterns, an Analysis of Variance (ANOVA) was conducted comparing mean changes in ideological positioning across the four media groups.

Table 2. ANOVA results for changes in ideological position across media groups.

Source of Variation	F-Statistics	p-Value	Effect Size (η2)
Between Groups	5.127	0.043	0.719

presents these results, revealing a statistically significant overall difference between groups (F = 5.127, p = 0.043). The substantial effect size (η² = 0.719) indicates that approximately 72% of the variance in ideological position changes is attributable to group differences. Follow-up post hoc comparisons with Bonferroni correction revealed specific significant differences between strongly conservative outlets and both moderately conservative (p = 0.037) and moderately progressive outlets (p = 0.014). Additionally, moderately progressive outlets differed significantly from strongly progressive outlets (p = 0.046).

These findings clearly support Hypothesis 1, demonstrating that polarization among media outlets was most pronounced prior to the wastewater release but moderated thereafter, particularly among moderate newspapers. Progressive and conservative outlets maintained or intensified their positions, whereas moderate outlets converged toward neutrality after implementation.

This empirical result aligns closely with theoretical frameworks explaining media dynamics during contentious policy debates. Specifically, Risk Perception Theory posits that heightened uncertainty before policy enactment intensifies ideological contestation [26], while the Issue–Attention Cycle anticipates decreased media intensity as speculative concerns give way to procedural reporting after risks fail to materialize [28]. Additionally, the observed shift toward neutrality among moderate outlets corresponds with normalization processes in environmental governance [30], indicating that media narratives become less polarized and increasingly oriented toward regulatory outcomes as policies move from debate to implementation.

5.2. Ideological Differences in Media Framing Strategies

The analysis of media framing reveals substantial differences in how progressive and conservative outlets constructed narratives about the Fukushima wastewater release. Figure 4 presents a word cloud analysis that illustrates the differential framing patterns employed by conservative and progressive media. The visualization contrasts the lexical emphases of the progressive Hankyoreh (blue text) and the conservative Chosun Ilbo (red text) in their coverage. These outlets were selected as representative cases based on their consistent ideological positioning throughout the period under analysis, as quantitatively validated by the Wordfish analysis.

Notably, Hankyoreh emphasizes terms such as “Government”, “Contaminated Water”, and “Fukushima”, which foreground concerns regarding environmental risks, public health, and governmental accountability. This lexical distribution aligns with progressive media’s emphasis on environmental risk assessment. In contrast, Chosun Ilbo prioritizes words such as “Representative”, “National Assembly”, “Democratic Party”, and “Conference”, thereby situating the issue within a political and institutional framework. This narrative strategy suggests that conservative outlets are more inclined to focus on political actors and legislative processes, thus diminishing the salience of environmental risks in favor of institutional governance and diplomatic considerations.

To quantify these framing differences, an environmental–scientific framing ratio was calculated for each outlet, measuring the relative emphasis on environmental risk narratives versus scientific validation frames.

Table 3. Environmental–scientific framing ratio by ideology and period.

Ideology	Period	Environment–Scientific Ratio	Environment Frame (%)	Scientific Frame (%)
Progressive	Before	0.748	6.58	2.21
Progressive	After	0.826	6.24	1.31
Conservative	Before	0.717	5.64	2.22
Conservative	After	0.813	5.71	1.32

presents these ratios across ideological groups before and after the wastewater release.

Bootstrap confidence interval analysis for the change in environmental–scientific ratio between progressive and conservative media produced a confidence interval of [−0.030, −0.006], which excludes zero. This provides statistical evidence of significant ideological differences in framing responses to the release event. The large effect size (Cohen’s d = 1.256) further indicates substantial practical significance in these framing differences.

Figure 5 shows these framing differences, illustrating how progressive media maintained consistently higher environmental-to-scientific ratios compared to conservative media both before and after the release, although both groups adjusted their framing following implementation.

These contrasting frames systematically contribute to a polarized media landscape, shaping public opinion as well as influencing policy debates and governmental responses. The analysis reveals distinct patterns in media framing that correspond to established ideological positions, which in turn reinforce ideological divisions in policy deliberation.

These empirical findings indicate that the Fukushima wastewater release exemplifies how media polarization shapes public perception and policy discourse. The analysis demonstrates that media outlets with opposing ideological leanings constructed fundamentally different narrative frameworks. Progressive outlets emphasized potential environmental and health risks, routinely scrutinizing government actions and advocating for more stringent safety measures. In contrast, conservative outlets prioritized geopolitical and institutional dimensions, highlighting diplomatic implications and governmental accountability rather than ecological concerns. These distinct narrative frameworks influence public understanding, contributing to opinion polarization and complicating consensus-building on policy responses.

The observed media polarization has significant implications for environmental governance. Divergent media coverage patterns can generate conflicting public demands, forcing policymakers to prioritize political considerations over evidence-based decision-making. Moreover, when scientific consensus is overshadowed by ideologically driven narratives, public trust in regulatory institutions may erode, weakening confidence in their ability to manage environmental risks effectively. Over time, such dynamics can undermine policy effectiveness, impeding critical interventions or fostering public resistance to policies perceived as ideologically motivated. The Fukushima case demonstrates the pressing need for balanced reporting and transparent communication to ensure that public discourse and policy decisions remain grounded in scientific evidence and serve the public interest.

6. Discussion and Conclusions

This study examines how media ideology shapes public discourse on environmental issues by analyzing coverage of the Fukushima wastewater release. The findings reveal that media outlets framed the issue through distinct ideological lenses, reinforcing differences in public interpretation. Even in policy areas traditionally viewed as technical or scientific, news media selectively emphasized certain aspects, contributing to ideological divisions. In South Korea, the Fukushima wastewater release became not only an environmental issue but also a politically contested subject, reflecting ideological debates over risk perception and regulatory trust.

Empirical analysis demonstrates that media polarization was strongest during the period of uncertainty before policy implementation, subsequently moderating as empirical evidence and procedural discussions gained prominence. This pattern of ideological convergence among moderate outlets was particularly notable, accounting for a substantial portion of the observed shifts in ideological positioning. Additionally, the analyses confirm systematic ideological differences in media framing, with progressive outlets highlighting environmental risks and governance shortcomings, while conservative outlets stressed scientific validation and regulatory legitimacy. The framing analysis reveals these consistent ideological differences in environmental versus scientific emphasis persisted even as both progressive and conservative outlets adjusted their coverage following implementation. The shift toward regulatory-focused coverage among moderate outlets after policy implementation further supports the argument that polarization decreases as policy debates transition into implementation phases.

Despite the insights provided by this study, several limitations should be acknowledged. First, this study focuses on media coverage without directly measuring shifts in public opinion. While integrating audience sentiment analysis could provide additional insights, such an approach requires a distinct dataset and methodological framework beyond the scope of this study. Future research could explore this relationship by combining media framing analysis with public sentiment data to assess how media narratives shape audience perceptions over time. Second, the study is limited to South Korea, and cross-national comparisons could provide an understanding of whether similar polarization patterns exist in other contexts. Examining how different political environments shape media framing of transboundary environmental issues would yield valuable insights. Third, while this study identifies a decline in media polarization following policy implementation, it does not empirically test the mechanisms driving this shift. Future research could examine regulatory, economic, or market-driven factors that may contribute to media de-escalation.

Methodologically, this study employs computational text analysis, specifically the Wordfish model, which allows for systematic measurement of ideological positioning across a large dataset. However, alternative computational techniques, such as AI-based methods involving natural language processing (NLP) or deep learning, could further validate differences in ideological framing by capturing nuanced lexical patterns and semantic structures. Further studies could benefit from integrating these advanced computational approaches or complementing them with qualitative analyses, such as discourse analysis or in-depth interviews, to achieve a more comprehensive understanding of media narratives and audience interpretations.

Nevertheless, this study contributes significantly to discussions on media polarization, environmental risk perception, and policy communication. The findings underscore the importance of strategic communication in environmental governance, particularly in navigating ideological divisions during policy implementation. The analysis demonstrates that ideological polarization intensifies in early stages of policy debates, yet media narratives evolve as policies move into implementation phases and empirical data replaces speculation. These findings align with established theories such as the Issue–Attention Cycle and Risk Perception Theory, showing how information availability fundamentally reshapes coverage patterns. Such temporal dynamics suggest that transparent and empirically grounded communication strategies could mitigate polarization and facilitate a more deliberative public discourse.

In conclusion, the Fukushima case demonstrates that media coverage of environmental issues is strongly influenced by ideological orientations. Empirical analyses reveal that ideological polarization intensified during periods of uncertainty prior to policy implementation but moderated afterward, particularly among moderate media outlets, as regulatory oversight and empirical assessments became dominant frames. Policymakers managing transboundary environmental risks should consider these patterns when developing communication strategies, particularly during the critical pre-implementation phase when media polarization peaks. The findings extend theoretical understanding of how media ideology functions across policy implementation cycles rather than remaining static. Such insights underscore the critical influence of media dynamics in shaping public discourse and emphasize the importance of strategic, evidence-based communication to effectively manage and reduce ideological polarization.