Understanding the Crisis
The decline of public-interest media is the product of several structural forces—economic, technological, and political—that interact differently across regions and contexts. Understanding these forces is essential to designing interventions that actually work.
The Collapse of the Advertising Model
For much of the twentieth century, journalism was sustained by a business model that had little to do with journalism itself. Newspapers and broadcasters sold access to audiences, and advertisers paid handsomely for it. The reporting effectively was, in economic terms, a byproduct, subsidized by the classified, car dealership, and department store ads.
The internet dismantled this arrangement almost overnight. Digital platforms offered advertisers something newspapers never could: precise targeting, vast scale, and measurable returns. Google and Meta captured the overwhelming majority of digital advertising growth, while news organizations—even those with large readerships—watched their share of ad revenue collapse. The money moved online to platforms that had no obligation to produce journalism and no incentive to support it.
We then end up in a situation where journalism is generating enormous public value but capturing almost none of it commercially. In a market organized around advertising, this discrepancy was hidden, with ads subsidizing the public-interest work. But once that subsidy disappeared, the mismatch was exposed.
While much of the research and philanthropic money has been focused on the decline of U.S. local journalism, it is far from a uniquely American story. Advertising-dependent models have contracted across Latin America, sub-Saharan Africa, South and Southeast Asia, and much of Europe. The details vary, but the underlying dynamic is global: digital platforms are much better at selling attention than news organizations, and no amount of digital transformation by newsrooms has been able to reverse this.
Platform Dominance & the Attention Economy
The collapse of the advertising model was intensified by the rise of platforms as the primary gatekeepers of information distribution.
Most people today encounter news not by visiting a newspaper’s homepage or tuning into a broadcast, but through their social media feeds, search results, and messaging apps. The visibility of a piece of journalism is determined by algorithms optimized for engagement, not for public interest.
Newsrooms produce content
Investigations, local reporting, accountability journalism
→
Platforms distribute it
Algorithmic feeds decide what gets seen, what gets buried
→
Platforms capture the value
Ad revenue, user data, and attention go to the platform, not the newsroom
This creates a dependency problem. News organizations rely on platforms for traffic and audience reach, but they have no control over the algorithms that determine distribution. A single change to Facebook’s News Feed or Google’s search algorithm can decimate a newsroom’s audience overnight.
Outlets that built their digital strategies around platform distribution have found themselves trapped: unable to reach audiences without the platforms, yet unable to achieve sustainability within them.
The latest exacerbation of this dynamic comes from the rise of generative artificial intelligence. AI trained on journalistic content can now produce summaries that compete directly with news articles. While these models rely on facts and reporting gathered by journalists, it is often without compensation or attribution. Users are rarely redirected to the original articles, hurting outlets who rely on ad or subscription revenue. As AI search tools replace traditional search traffic, news organizations risk losing yet another channel through which audiences once found their work.
Political Pressure & Capture
Market forces are not the only threat to public-interest media. In many parts of the world, journalism is undermined not by commercial failure but by deliberate political action.
The spectrum of political interference in media
Political pressure on media is not confined to authoritarian contexts. In established democracies, subtler mechanisms can achieve similar effects. These pressures may be less dramatic than jailing a journalist, but they erode the conditions for independent reporting just as effectively over time.
The key insight is that political interference exists on a spectrum, and many democracies sit closer to the middle than their self-image would suggest. Media independence depends not just on the absence of overt censorship but on the presence of institutional safeguards that insulate journalism from political control. Where those safeguards are weak, even well-intentioned governments can distort the media landscape.
The Uneven Geography of the Crisis
Not all media ecosystems face the same problems in the same ways. Any attempt to address the crisis must account for this variation.
Local vs. national
Local journalism is most vulnerable
Thin margins, smaller audiences, fewer revenue alternatives. Yet this is where democratic stakes are highest: school boards, municipal budgets, local courts.
High-income vs. low-income
The Global South faces compounding pressure
Weaker markets, less philanthropy, greater political risk, less international attention. Every structural force hits harder with fewer buffers.
English vs. non-English
Funding concentrates in English-language markets
Research, platform attention, and philanthropic support skew heavily toward English. Indigenous-language and minority-language media face compounding barriers.
Exported solutions
Western frameworks don't always translate
Tools and funding mechanisms developed in Paris or the Silicon Valley may not fit a community radio station in rural Senegal or a newsletter in the Philippines.
It’s important to understand the implications of these differences. A publication like the New York Times will have a very different experience negotiating with Meta or Google compared to a small rural outlet. Different contexts require different interventions, and solutions designed for well-resourced democracies often fail when exported to environments where the challenges are fundamentally different.
