Making Media Work for Democracy: Lessons from the Past, a Vision for Our Future

Our democratic system is experiencing extreme strain. The checks and balances that have long defined our government are eroding, partisan polarization is increasing, and policymakers are failing to meet voters’ needs. In a Pew Research Center survey, 69 percent of Americans expressed dissatisfaction with the way democracy is working in the United States.

Of the many factors that contribute to this condition, one stands out: the media. The past twenty-odd years have seen a vast transformation in the systems Americans use to inform, entertain, and engage with one another. We have gone from interacting primarily in person, supplemented by comparatively primitive communication devices, to engaging largely over globalized social networks. Our primary information sources have shifted from shared print and television news to social media feeds tailored to sustain our engagement. Even our entertainment is fragmented and dished out in an ever-expanding buffet of content.

For nearly two decades, up to 29 million Americans tuned in nightly to watch Walter Cronkite, while nearly 106 million Americans viewed the finale of M*A*S*H. The broadcast networks wielded immense gatekeeping power, often dictating a narrow picture of American society. But they also created common stories, characters, and events, allowing millions of Americans to enjoy a shared experience. Those moments are largely gone. 

Digital media brought its own democratic gains, allowing voices and communities once excluded to reach enormous audiences, connect people across geography, and lower the cost of civic organization. But those gains came with a price: a loss of commonality and the rise of new forms of control over what information people consume. For democracy to thrive, we must preserve the gains made by digital media while rebuilding commonality, agency, and human understanding.

Media and Self-Governance in America

Democratic renewal has often accompanied changes in how citizens communicated with one another. The founders debated the new republic through pamphlets, newspapers, and books. During the struggle against slavery, abolitionist newspapers and slave narratives expanded readers’ moral horizons. Later, radio allowed Franklin Roosevelt to speak directly to citizens in their homes, while television brought the violence of segregation in Birmingham to a national audience.

None of these communication systems was inherently democratic, nor did Americans adapt to them without difficulty. The democratic value of these various channels emerged through the institutions, norms, and social purposes constructed around them, as Americans learned to fit new forms of communication into shared civic life. The challenge before us is to do the same with a communications system that is more decentralized, personalized, and powerful than ever before.

Pluralism in the Algorithmic Age

Social media is no longer primarily social. Increasingly, it functions as algorithmic media, where users encounter not only what they choose to follow, but what a recommendation system predicts will hold their attention. By repeatedly serving content shown to perform well, these systems can narrow the range of sources, subjects, and perspectives users encounter. One promising response is algorithmic pluralism, drawing in part on Francis Fukuyama’s proposal for social media “middleware.” Under this model, users no longer depend solely on a dominant platform’s recommendation system, instead choosing among third-party services that rank and curate content according to different values and purposes. A newspaper’s model might emphasize reported news, a civic institution might prioritize local information, and another provider might focus on friends, cultural discovery, or constructive engagement across partisan differences.

Emphasis on engagement did not come from social media companies being indifferent to democratic life. It stems from a highly concentrated market for attention in which engagement is the clearest source of value, pushing platforms to make their products evermore responsive, personalized, and compelling. Middleware could introduce institutions with different incentives into that market, including newspapers, civic organizations, nonprofits, and other providers that compete on trust, local information, source diversity, privacy, or the quality of the experience they create.

Rather than dictating what constitutes a healthy feed, the government’s role should be limited to establishing the conditions in which such a market can emerge. Congress could require the largest platforms to maintain interoperable Application Programming Interfaces (APIs)—a set of rules that lets different software programs talk to each other—through which qualified middleware providers, with a user’s permission, could access the information necessary to construct an alternative feed. A federal regulator could set privacy and transparency standards for that access, prevent platforms from disadvantaging competing services, and ensure researchers can evaluate how different recommendation systems shape what users encounter, all without dictating what any of those systems should prioritize.

From a technological perspective, much of this design is possible. Modern recommendation systems routinely balance several objectives at once, and Google researchers have described a large-scale video ranking system built to optimize multiple competing goals. A 2025 experiment reranked participants’ actual feeds on X, changing their exposure to posts expressing partisan animosity and antidemocratic attitudes, and found corresponding changes in affective polarization. These studies cannot define which democratic values should guide recommendation systems, but they do demonstrate that engagement does not need to be their sole objective.

How Media Brings Us Together

As partisan polarization becomes more social and affective, political disagreement increasingly shapes how Americans judge one another’s character, making pluralistic compromise and the acceptance of democratic outcomes more difficult. Pew found that 72 percent of Republicans and 63 percent of Democrats view members of the other party as more immoral than other Americans, while More in Common found that partisans substantially exaggerate the extremity of the other party’s views.

Through my work at More Like US, I’ve explored how variations of intergroup contact theory, especially vicarious and parasocial contact, offer one path toward correcting these misperceptions through media. These two forms of contact differ mechanistically. Vicarious contact occurs when someone watches a person similar to them engage positively, or simply normally, with a member of a group toward which they may hold some prejudice, reducing that prejudice through observation. Parasocial contact develops through a one-sided relationship with a character or creator from that group, gradually challenging the viewer’s assumptions and misperceptions. These approaches have been shown to reduce prejudice across society, from attitudes toward autistic people to immigrant children, and more recently people with differing partisan affiliation. In comparison with other forms of depolarization, such as in-person, cross-partisan workshops, media-based contact could reduce the cost per participant by roughly 94 percent while retaining about 76 percent of the workshop’s effect on affective polarization, making it substantially more scalable.

Entertainment media offers a particularly promising way to apply these mechanisms. Sitcoms, dramas, or other digital content can portray partisan differences inside ordinary relationships, with disagreement existing alongside humor, affection, frustration, and responsibility for one another. Using clips from The Conners, researchers found that positive contact between characters affiliated with different parties could lessen partisan tension, especially when viewers formed parasocial relationships with the characters and experienced them crossing partisan divides. Within the ever-growing short-form social media ecosystem, these strategies could appear through comedy, personal stories, or brief scenes showing people with different views engaging meaningfully and candidly across those divides. 

The significance of this content goes beyond reducing partisan tension. Before Americans can meaningfully share a culture or political community, they need a baseline belief that those on the other side are not enemies, but people whose lives, relationships, frustrations, and convictions are also human. Media cannot erase political disagreement, nor should it, but it can help restore that recognition, allowing curiosity, restraint, and relationships that survive disagreement to enter content people consume voluntarily rather than arriving as a formal civics lesson.

Rebuilding Shared Culture

Alongside greater user choice and more constructive forms of mediated contact, we must acknowledge the decline of shared cultural moments. A democratic society benefits when millions watch the same story, event, or performance, knowing that fellow citizens have experienced it as well. Sports, television, music, and comedy can create this commonality more naturally than civic programming, building forms of knowledge, affection, humor, and identification that extend beyond elections and partisan identity. 

Rebuilding shared culture is not a regulatory project. Government should not decide what Americans watch, read, or listen to. Nor has mass cultural attention entirely faded. Stranger Things generated 40 billion minutes of U.S. viewing in 2025, while the 2026 Super Bowl drew 125.6 million viewers. 

The challenge lies in the increasing separation of audiences across platforms, subscriptions, and personalized environments, making it difficult for even widely popular work to become common. When cultural events break through those divisions, we should make them as broadly accessible as possible. Britain and Australia protect broad, free access to events of national importance. There is no obvious equivalent for ordinary entertainment, but media companies, philanthropies, and cultural institutions could experiment with free-access windows or broader distribution when a program enters the national conversation.

Media reforms cannot repair every weakness in American democracy, but they can help build more effective institutions, inform voters, and promote the election of more responsive leaders. Democratic renewal depends on how citizens communicate with, encounter, and imagine one another. By fixing how Americans see and hear each other, we can create the conditions for democracy to thrive.