Media does not just report on politics; it actively shapes what people think about, how they feel about candidates and issues, and even whether they vote at all. The relationship runs in both directions, too, with politicians and campaigns increasingly engineering content designed to exploit the way modern media ecosystems work. From agenda-setting on network television to deepfakes on social media, the entanglement of media and politics touches nearly every aspect of democratic life, often in ways that are invisible to the audience consuming it.
How Media Decides What You Think About
The oldest and most well-documented power of media in politics is not telling you what to think but telling you what to think about. When news outlets devote sustained attention to an issue, that issue climbs in public importance. Research on same-sex marriage coverage, for instance, found that the emotional tone of how news outlets framed the issue was more strongly associated with shifts in public opinion than the specific policy arguments presented. Coverage shaped attitudes not mainly through factual persuasion but through the feelings it attached to the topic. And the effects were not fleeting: on contested issues, media influence on public opinion appeared to accumulate over the long term rather than spike and fade.1Newspaper Research Journal. What changed public opinion on the same-sex marriage issue? New implications of attribute measures and attribute priming in media agenda setting
This matters because the sheer volume of potential political topics at any moment far exceeds what news organizations can cover. The choices editors and algorithms make about which stories get prominence, and how those stories are framed emotionally, function as a kind of invisible filter on democratic debate. If an issue never gets sustained coverage, it rarely registers as a public priority, regardless of how many people it affects.
Why Everyone Thinks the Media Is Biased Against Their Side
One of the more counterintuitive findings in this field is that partisans on opposite sides of an issue routinely watch the same news segment and both come away convinced it was slanted against them. This pattern, known as the hostile media effect, was first documented in the 1980s using coverage of the Beirut massacre. Pro-Israeli and pro-Arab viewers watched identical news footage and each group reported that the coverage contained more negative references to their own side. Each group also predicted the segment would push neutral viewers toward the opposing position.2PubMed. The hostile media phenomenon: biased perception and perceptions of media bias in coverage of the Beirut massacre
Two things drive this. First, people with strong opinions have a different mental baseline for what “balanced” looks like. If you believe your side’s case is overwhelmingly strong, any coverage that gives equal time to the opposition feels like a distortion. Second, partisans tend to remember and weight the parts of coverage that feel threatening to their position more heavily than the parts that support it. Prior beliefs about whether the media is biased in general also amplify the effect, creating a feedback loop where each new piece of coverage reinforces the conviction that the media cannot be trusted.3Journal of Experimental Social Psychology. The Causes of Hostile Media Judgments
There is a potential antidote, though. When people encounter political content in a context that includes a mix of viewpoints, rather than a feed dominated by one side, the perceived bias gap between friendly and unfriendly sources shrinks. Exposure to opinion diversity can also reduce hostility toward people on the other side.4Group Processes & Intergroup Relations. The role of opinion diversity in the hostile media effect: Partisan exposure on social media, affective polarization, and political participation
Confirmation Bias and the Echo Chamber Problem
People do not consume media at random. They gravitate toward sources that confirm what they already believe, and this tendency has real consequences for how political information travels. A field experiment studying populist partisans found that people with high political commitment, measured by voting intention, showed classic confirmation bias: they clicked on news stories that aligned with their existing views and avoided stories that challenged them. Interestingly, people with high affinity for a party but lower commitment behaved differently, showing more willingness to engage with opposing viewpoints, especially when browsing rather than clicking through headlines.5Media and Communication. Partisan Selective Exposure in Times of Political and Technological Upheaval: A Social Media Field Experiment
Social media platforms amplify this tendency by design. Analysis of user behavior on Facebook and Twitter found that people cluster into homogeneous groups that share and reinforce a common narrative. Facebook showed higher levels of information segregation than Reddit, suggesting that platform architecture plays a significant role in how sealed-off these bubbles become.6PubMed Central. The echo chamber effect on social media Within these clusters, content spreads along predictable pathways, and each echo chamber develops its own patterns of information cascading. The homogeneity of a cluster, not the truthfulness of the content, is the primary driver of how far a piece of information travels.7PubMed Central. The spreading of misinformation online
The practical upshot is that two people living in the same city, working at the same company, can inhabit completely different informational worlds. One sees a political scandal as the top story of the week; the other has never heard of it. This fragmentation makes shared political reality harder to maintain.
How Partisan Media Fuels Emotional Polarization
The gap between left and right in many democracies is increasingly defined not by policy disagreement but by dislike and distrust of the other side. Partisan media plays a measurable role in widening this emotional divide. Panel analysis has shown that exposure to pro-party television sources strengthens affective polarization among viewers, and the mechanism runs through discrete negative emotions: watching partisan media makes people angrier at and more fearful of opposing candidates, and those emotions in turn drive the wedge deeper between political camps.8Journalism & Mass Communication Quarterly. Partisan Information Sources and Affective Polarization: Panel Analysis of the Mediating Role of Anger and Fear
This is not just a story about cable news. The same dynamics play out on social media, where algorithmic feeds can serve a steady diet of outrage-inducing content from one political direction. Filter bubbles limit exposure to opposing views and reinforce confirmation bias, while echo chambers encourage interaction only with like-minded users, compounding polarization over time.9Dinamika Penelitian: Media Komunikasi Penelitian Sosial Keagamaan. The Role of AI, Filter Bubbles, and Echo Chambers in Political and Religious Polarization on Social Media
Misinformation Spreads Easily, But Corrections Actually Work
A common source of despair is the feeling that once false information enters the political bloodstream, nothing can get it out. The evidence is more hopeful than that. Simultaneous experiments across four countries, Argentina, Nigeria, South Africa, and the United Kingdom, found that fact-checking reduced belief in misinformation by a meaningful amount on a five-point accuracy scale, and most of those effects persisted for more than two weeks. Meanwhile, mere exposure to a piece of misinformation only nudged false beliefs upward by a tiny fraction on the same scale. In other words, corrections punch harder than the lies they are correcting.10PubMed Central. The global effectiveness of fact-checking: Evidence from simultaneous experiments in Argentina, Nigeria, South Africa, and the United Kingdom
There is a catch, though. People who are highly confident in their own political knowledge are harder to reach with corrections. Research on inoculation strategies, where people are pre-emptively warned about common false claims, found that these treatments reduced support for misperceptions in general but were ineffective among the politically overconfident.11APSA Preprints Archive. Political Overconfidence Inhibits the Effectiveness of Inoculations Against Misperceptions The people who most need correcting are often the least receptive to it, which partly explains why fact-checking alone cannot solve the problem of political misinformation.
Deepfakes and the Erosion of Trust
AI-generated synthetic media has added a new dimension to political manipulation. Research consistently finds that people struggle to identify deepfake videos, and their opinions shift after watching them.12Journal of Creative Communications. Artificial Intelligence and Political Deepfakes: Shaping Citizen Perceptions Through Misinformation But the damage goes beyond any single deception. A systematic review of the psychological effects of political deepfakes found that even when viewers did not fully believe the manipulated content, their trust in news, political information generally, and political institutions declined after exposure. The mere existence of deepfakes makes people more skeptical of everything, including things that are true.13Computers in Human Behavior Reports. Exploring how political deepfakes trigger cognitive processes and downstream psychological and behavioral outcomes: A systematic review
One surprising finding complicates the expected story about polarization. An experimental study found that exposure to news about an opposing party’s use of deepfakes did not actually increase polarization toward parties, coalitions, or leaders. Approval for candidates’ use of AI tools was negatively associated with polarization, suggesting that when people recognize AI as a campaign tactic, it may generate backlash rather than the intended persuasive effect.14Social Science Computer Review. Campaign Deepfakes and Affective Polarization: The Role of Artificial Intelligence in Campaigns in Shaping Voter Attitudes
Bots and Fake Grassroots Campaigns
Political astroturfing, where coordinated fake accounts simulate grassroots political activity online, is a global phenomenon. Analysis of known astroturfing campaigns found that coordinated accounts could be detected with high accuracy: on average, about three-quarters of astroturfing accounts were identified through their unusually synchronized posting behavior. In one well-studied case, an Internet Research Agency campaign in the United States, more than 80 percent of the astroturfing accounts were detectable, with a false-positive rate hovering around just one percent.15PubMed Central. Coordination patterns reveal online political astroturfing across the world
What do these bots actually accomplish? Research on how humans respond to bot interactions found that politically oriented astroturf bots provoke engagement: people who interacted with them posted more, not less. But regardless of the type of bot, direct interaction lowered the average emotional tone of human users’ posts. Bots drive conversation, but they make it angrier.16Scientific Reports. Social bots spoil activist sentiment without eroding engagement
When Local News Disappears
While much of the attention in media and politics focuses on national outlets and social platforms, some of the most consequential effects involve local journalism. An analysis matching local newspapers in California to the cities they covered found that municipalities served by papers with sharp declines in newsroom staffing experienced significantly reduced political competition in mayoral races. There was also suggestive evidence of lower voter turnout in those areas.17Urban Affairs Review. Political Consequences of the Endangered Local Watchdog: Newspaper Decline and Mayoral Elections in the United States
Canadian data tells a similar story. Voter turnout in municipal elections was positively correlated with newspaper health, measured by both the number of papers per population and how frequently the largest local paper published.18Canadian Journal of Political Science. Does Democracy Die in Darkness? An Examination of the Relationship between Local Newspaper Health and Turnout in Municipal Politics Without a local watchdog, citizens know less about what their government is doing, incumbents face less scrutiny, and fewer people bother to vote. The decline of local news is not just an industry problem; it is a democratic one.
Microtargeting and Personalized Political Persuasion
Political campaigns now routinely use data to tailor messages to specific audience segments. Research testing microtargeting strategies, combining message pretesting with machine learning, found that targeted messages outperformed untargeted ones by roughly 70 percent or more when trying to shift attitudes on a single policy issue. Notably, though, targeting by more than one variable at a time did not yield additional gains, and the advantage depended heavily on context. For one of the two policy issues studied, microtargeting barely outperformed simpler approaches.19PubMed Central. Quantifying the potential persuasive returns to political microtargeting
The addition of generative AI to this toolkit raises the stakes further. Experiments matching political ads to individual personality profiles found that better-matched messages were rated as more persuasive. The larger the mismatch between the ad’s style and the viewer’s personality, the less persuasive it felt.20PubMed Central. The persuasive effects of political microtargeting in the age of generative artificial intelligence AI can now generate thousands of message variants cheaply, making personality-matched targeting scalable in a way it was not even a few years ago. The power of microtargeting is real but uneven, and its effectiveness depends as much on the issue and the audience as on the sophistication of the technology.
The Knowledge Gap That More Choice Creates
The explosion of media options over the past two decades was supposed to create a better-informed citizenry. In some ways it has done the opposite. A foundational study of the high-choice media environment found that people who preferred news consumed more of it and became more politically knowledgeable and more likely to vote. But people who preferred entertainment simply abandoned news altogether, becoming less informed and less likely to show up at the polls. As media choice expanded, the gap in political knowledge and turnout between these two groups widened.21American Journal of Political Science. News vs. Entertainment: How Increasing Media Choice Widens Gaps in Political Knowledge and Turnout
More recent experimental work suggests the problem goes deeper than motivation. Even when interventions equalized people’s information-seeking behavior, knowledge gains were concentrated among participants with higher education or stronger baseline civic knowledge. These individuals appeared more effective at navigating search results and extracting useful information. Closing the knowledge gap requires more than just putting information in front of people; it requires building the skills to use it.22arXiv. The Knowledge Gap in a High-Choice Media Environment: Experimental Evidence from Online Search
Countries with strong public service broadcasting systems fare somewhat better on these metrics. Cross-national comparisons have found that populations in countries with robust public broadcasters tend to be better informed about government, more trusting, and more likely to participate in democratic politics than populations in countries dominated by commercial broadcasting.23The Political Quarterly. Public Service and Commercial Broadcasting: Impacts on Politics and Society Public broadcasting does not eliminate knowledge gaps, but it appears to narrow them by keeping a baseline of political information in front of audiences who might otherwise tune out entirely.
Social Media as a Tool for Political Movements
The relationship between media and politics is not purely top-down. Social media has given ordinary people new tools to organize, coordinate, and draw attention to political causes. Studies of protest movements in the United States, Spain, Turkey, and Ukraine have documented how platforms facilitate the rapid exchange of logistical information, such as transportation routes, police locations, and legal support contacts, that would have been far harder to distribute before. Platforms also serve as vehicles for emotional and motivational content: messages emphasizing anger, group solidarity, fairness, and ideological themes circulate alongside the practical details and help sustain participation.24Political Psychology. How Social Media Facilitates Political Protest: Information, Motivation, and Social Networks
The structural features of online networks also shape which movements succeed and which fizzle. Networks with certain connectivity patterns are better at spreading information broadly, while others create tight clusters that are good for internal coordination but poor at reaching new audiences. Political ideology can influence these structures, meaning that movements on different sides of the political spectrum may face different organizational advantages and constraints online.
Media Literacy as a Practical Skill
Given everything above, the ability to evaluate media critically is not a luxury; it is a basic requirement for navigating political life. A digital media literacy intervention tested in both the United States and India found that a brief training improved people’s ability to distinguish between mainstream and false news headlines by about 27 percent in a nationally representative U.S. sample and about 18 percent in an educated Indian sample. In the U.S. group, the improvement was still measurable several weeks later. However, the intervention had no effect in a largely rural area of northern India where social media use was much lower, suggesting that literacy interventions work best when people are actually encountering the problem they are being trained to recognize.25PubMed Central. A digital media literacy intervention increases discernment between mainstream and false news in the United States and India
The intervention worked by making participants more skeptical across the board, reducing perceived accuracy of both real and false headlines. But the reduction was larger for false headlines, meaning net discernment improved. This is an important nuance: the goal is not to make people trust everything less, but to create a gap between how they evaluate credible and non-credible sources. The fact that this gap can be widened with relatively modest training is one of the more encouraging findings in a field that often feels dominated by bad news.
Platform Regulation and the Free Speech Tension
Governments have begun trying to regulate the online spaces where much of political media now lives, but these efforts generate their own political dynamics. Germany’s NetzDG law, which required platforms to remove certain categories of illegal content, sparked intense public debate that itself became a political flashpoint. Analysis of the Twitter discussion around the law revealed two distinct politically aligned clusters: one connected to the right-wing Alternative für Deutschland party that opposed the law as censorship, and one linked to the liberal Free Democratic Party. A third cluster consisted of media and technology stakeholders. “Alternative” right-wing media outlets and specialized tech websites played outsized roles in some clusters, while legal experts held central positions across all groups.26Policy & Internet. NetzDG on Twitter: A Social Network Analysis of Stakeholders in Public Discourse About Platform Regulation in Germany
The German experience illustrates a recurring tension. Regulation aimed at cleaning up political media environments inevitably becomes political itself, with different factions framing the same law as either democratic protection or authoritarian overreach. The stakeholders with the loudest voices in those debates are often not ordinary users but media organizations and political operatives with their own interests in how the rules get written. Any regulatory approach to media and politics has to contend with the fact that the regulation itself will be filtered through the same partisan dynamics it is trying to address.