Since 2005, the United States has lost roughly a third of its newspapers — more than 3,000 outlets — and papers keep closing at an average pace of about 2.5 per week, per Northwestern University's Medill School of Journalism, Local News Initiative reporting in its State of Local News reports, with the 2024 edition the latest cited here. More than 200 counties now have no local news outlet at all. Those are the documented numbers, and they carry a narrower lesson than either the eulogies or the pep talks: the decline is concentrated, uneven, and — in the places it hits — measurable in civic terms.
The Medill reports are the standard citation in this field because their method is published: they track outlet closures, launches and ownership changes county by county. What follows reads the reports as a case study in what transfers and what doesn't for practitioners deciding where to work and what to build.
Where did the closures actually concentrate?
In small communities, and in papers already acquired by investment-backed chains. The Medill 2024 reporting associates closure risk with two documented factors: market size (communities under about 5,000 population dominate the loss column) and ownership type, with private-equity- and hedge-fund-owned papers closing at higher rates than independent or family-owned ones. The pattern is not uniform decline — it is thinning at the bottom of the market while metro distress makes the headlines.
That distinction matters for anyone planning a career or a startup. A metro daily's problems are revenue-model problems; a five-thousand-person town's problems are total-market-size problems. The same word — "unsustainable" — describes two different diseases.
What did the growth side of the ledger show?
Philanthropy-funded and digital-only startups, at a pace that has not kept up with closures. The same Medill reporting counts hundreds of new digital local outlets launched in the period, disproportionately in metro and suburban markets rather than the rural counties losing coverage — so the replacement activity clusters where the audience density supports it. Counties that gained a newsroom rarely overlapped with counties that lost one.
Funder disclosure, every time, per this site's standing rule: the Medill State of Local News reports are supported by journalism-philanthropy funding, including the MacArthur Foundation and others named on the reports' own pages. That funding does not change the arithmetic, and readers should know it exists.
What transfers from the numbers — and what doesn't?
The spine of the case, in the fixed order:
- What the field tried: the traditional advertising bundle, then paywalls, then philanthropy and nonprofit models, documented across the Medill series and the Pew Research Center's journalism fact sheets.
- The documented numbers: roughly a third of papers lost since 2005; about 2.5 closures per week; more than 200 counties with no outlet, per Medill 2024.
- What transferred: digital startups with audience density and philanthropic backing have survived in metros; that much the launch-and-survival data show.
- What didn't, and why: the smallest markets, where no advertising, subscription or donor base clears the cost of even one journalist — the reports' structural point, not a judgment on the people trying.
What does this mean for practitioners?
Job-market realism and target-market realism, in one number each. Pew Research Center's newspaper industry fact sheets, built on Alliance for Audited Media and industry data, show circulation falling from tens of millions of weekday copies in the mid-2000s to a fraction of that — so the employment base of the profession has shrunk with it. The Medill data's practitioner lesson: coverage deserts are also competition deserts, and several documented nonprofit launches have succeeded specifically in counties other outlets abandoned. That worked for a newsroom of six with a foundation grant. It says nothing about a metro daily, and the reports themselves flag the difference.
What the numbers don't settle
Causation at the community level. Medill and independent researchers have associated news-desert counties with lower civic participation — the reports pair desert counties with lower turnout and higher government borrowing costs where those associations have been studied — but the data is ecological, tracking counties, not individuals. The honest close: the shrinkage is documented, the geography is documented, and the civic consequences are documented associations, not proven mechanisms. For a field that traffics in precision, keeping those straight is the craft lesson of the whole dataset.
For more context, read What a News Desert Actually Measures — and What It Misses.