Comparative researchers keep finding the same puzzle. Take citizens in Finland and the United States, match them for education and interest in politics, and ask them basic questions about international affairs. The Finns tend to answer more questions correctly. More strikingly, the gap between their most and least educated citizens is smaller.

Explanations based on individuals don't account for this. Americans are not less curious, and Finns are not born better informed. The difference lies upstream, in how each country's media system is built: who owns the outlets, how they earn money, and what rules govern them.

This is the core claim of structural media analysis. Political knowledge is not simply a product of individual effort. It is an institutional output, shaped by incentive structures most citizens never see. Mapping those structures shows why some societies share a common factual baseline while others split into separate informational worlds. It also shows who now holds the power to decide which way a society goes.

Every news organisation answers to whoever pays its bills, and different payers create different incentives. A commercial broadcaster funded by advertising must deliver audiences to advertisers. A public broadcaster funded by licence fees or taxation must justify itself to legislators and, ultimately, to the public as a whole. Neither organisation needs explicit instructions. The funding relationship quietly defines what counts as success.

These pressures shape content in consistent ways. Commercial models reward programming that maximises audience share at minimum cost. That tends to favour entertainment, crime and human-interest stories over sustained coverage of policy or foreign affairs. Public service models, especially those with legal mandates for universal service, are pushed the other way. They schedule news at peak hours, cover unglamorous subjects, and serve audiences that advertisers undervalue.

Comparative research by James Curran, Shanto Iyengar and colleagues found that market-dominated systems tended to produce less hard news and wider knowledge gaps between advantaged and disadvantaged groups. Public service systems did not turn everyone into an expert, but they raised the floor. Part of the effect came from inadvertent exposure: people who switched on the television for something else ran into news anyway.

Neither model is free of distortion. Public broadcasters risk political capture, particularly where governing parties control appointments and budgets. That pattern is visible in several European democracies. Commercial outlets can be more independent of the state precisely because they depend on the market instead. So the structural question is not which model is pure. It is which dependencies a society is willing to accept, and which safeguards will buffer them. Arm's-length boards and multi-year funding settlements are two common examples.

Takeaway

Ask who pays for the news you consume, and you will usually see the outline of what it covers and what it leaves out. Funding is editorial policy by other means.

For much of the twentieth century, television viewers in many countries had only a handful of channels. When the news was on, much of the audience watched it, whether they cared about politics or not. Political scientist Markus Prior called this an inadvertent audience: people who learned about public affairs as a side effect of having little else to watch.

Cable, satellite and later the internet dissolved that captive audience. Expanding choice did not reduce the supply of political information. It vastly increased it. What changed was who consumed it. People who liked news consumed more than ever, while people who preferred entertainment could now avoid news entirely. Prior linked this shift to a widening knowledge gap and to more polarised electorates. Less interested citizens drifted away from voting, while the most engaged, who were often the most partisan, stayed.

Fragmentation also erodes something subtler: common knowledge. In game-theoretic terms, common knowledge is not just what everyone knows. It is what everyone knows that everyone else knows. A shared evening bulletin created that layer, so citizens could assume a baseline of facts when they argued, coordinated or held officials to account. When audiences split into separate streams, that assumption weakens.

The consequences for institutions are significant. Politicians increasingly address segmented audiences rather than a general public. Accountability that depended on everyone paying attention at once loses force, such as a scandal the whole country heard about on the same night. In market terms, choice is usually treated as an unambiguous good. In political terms, its effects are distributional. It rewards those who are already engaged and quietly releases everyone else from paying attention.

Takeaway

More information does not necessarily produce a better-informed public; it can produce a more unequally informed one. Abundance changes who learns, not just how much is available to learn.

Agenda-setting research goes back to Maxwell McCombs and Donald Shaw's study of the 1968 US presidential campaign. It established a durable insight: news media may not tell people what to think, but they are remarkably effective at telling them what to think about. For decades, that power sat with editors who decided what went on the front page or led the bulletin.

Much of that power has since migrated. Many people now get their news through search engines, social feeds and video platforms, whose ranking systems decide which stories surface. These intermediaries rarely produce journalism, yet they shape its visibility and its economics. Increasingly they shape its form too, as publishers adapt headlines and formats to whatever the algorithm rewards.

This kind of power is distinctive. Editorial judgements were made by identifiable people working under professional norms. In many countries they were also bound by regulatory obligations, such as impartiality rules for broadcasters. Platform curation works differently in three ways:
• It is spread across ranking signals optimised mostly for engagement.
• It is governed by terms of service rather than public law.
• It can be changed overnight.
When a platform deprioritises news in its feed, traffic to publishers can collapse with no legislative debate and little public visibility.

Regulators have begun responding. Some jurisdictions now have bargaining codes that require platforms to negotiate with publishers. Others impose transparency obligations for algorithmic systems or rules on systemic risk for the largest services. But these measures follow a familiar incremental pattern: they adjust the edges of an arrangement whose core logic remains commercial. The gatekeeper has changed, and the institutions designed to hold gatekeepers accountable are still catching up.

Takeaway

Many of today's most consequential editorial decisions are made by systems that do not see themselves as editors. Power that denies it exists is the hardest kind to hold accountable.

Taken together, these three dynamics point to a single pattern. What citizens know about politics is largely determined by arrangements they did not choose and rarely notice: funding models, the amount of choice available, and the ranking logic of intermediaries.

Each layer distributes knowledge unevenly. Public service structures tend to narrow knowledge gaps, and fragmented markets tend to widen them. Platform curation adds a gatekeeper whose priorities are commercial and largely unaccountable to public institutions.

For anyone trying to understand governance, the implication is that information is itself a policy outcome. Debates about broadcasting charters, competition law and platform regulation are not side issues for media specialists. They are decisions about who in a democracy gets to know what, and who decides.