Gannett, publisher of USA Today and more than two hundred local newspapers, had about 1.45 million paid digital subscribers at the end of the third quarter of 2025. That works out to roughly seven thousand per paper. The number is not holding steady. It is down twenty-six percent year over year.
That is the state of paywalled local news in America at the largest scale anyone has tried it. A company with an enormous portfolio of daily papers, a national flagship, years of subscription infrastructure, and professional marketing teams is collecting an average audience smaller than a high school enrollment per title — and losing that audience faster than it is replacing it. Everything that follows is an attempt to explain why.
What the National Paywalls Prove, and What They Don't
The New York Times has more than twelve million paid subscribers. The Wall Street Journal has more than four million. Both paywalls work. Both are routinely cited as the model a struggling local paper should follow.
Neither example is relevant.
What the Times and the Journal have in common is not a clever paywall design. It is the kind of national reach and authority that took decades to build and has been reinforced by every editorial decision since. These publications assembled audiences — people who had a reason to pay for the product before the product was behind a gate. The wall came after the trust. The paywall is an infrastructure for monetizing attention that already exists at scale.
A local paper in the year 2026 has the opposite problem. It does not yet have the attention it is trying to monetize. Its readership has been shrinking for twenty years. The remaining audience is older, more loyal, and more likely to be renewing a subscription that started in print. Any growth has to come from a new reader — someone who clicks a link, lands on an unfamiliar publication, and decides whether to come back. A paywall is friction at the exact moment that decision is being made.
The research on this is consistent. The American Press Institute's long-running work on subscription conversion finds that digital paid-audience growth comes from the most engaged readers — roughly five to ten percent of them. That is five to ten percent not of an audience, but of a narrow subset of an audience that was already showing up regularly. A struggling local paper does not have that subset in sufficient size. That is the point.
What the Paywall Trades
The strongest case for paywalled local news is not the one most publishers use in marketing copy. It is the case Richard Tofel, Dan Kennedy, and a handful of other thoughtful people have made: reader revenue is the only revenue that makes a newsroom independent of advertisers and platform decisions. A paper that lives on subscriptions does not have to care what Google thinks of its headlines or what an advertiser thinks of its coverage.
That argument is real. It is also insufficient. Independence requires readers. A local paywall that converts a few thousand of the most-engaged members of an already-declining audience does not produce independence. It produces a private subscription product for a small loyal club, dressed in the language of public-interest journalism. The independence is theoretical because the scale is missing. You cannot be independent from the advertiser market if the alternative is being irrelevant.
The trade a local paywall makes is this: you keep the revenue from the readers who were already staying. You lose any chance of building a new audience that could have replaced the one that is leaving.
The Forecast Is Not a Forecast Anymore
The other reason paywalled local news is collapsing into itself is harder to quantify but easier to see. As of early 2026, ChatGPT has about 900 million weekly users. Perplexity is growing quickly. Google's AI Overviews are now the default answer on a large share of news-adjacent queries. The behavior of opening a browser and clicking through to a publication is being replaced — at scale, already — by asking an agent and reading a summary.
That shift does not reward publications with hard paywalls. Agents do not subscribe. They index what they can reach. The publications that become the sources named in an AI-curated morning briefing will be the ones an agent could reach, and keep reaching, without running into a subscription form. For most of the last twenty years, local news lost audience to Facebook. For the next five, it will lose it to whichever agent a reader starts trusting most. The paywall is a bet that the old pathway is coming back. It isn't.
What the Trade Actually Costs
A paywall on a declining local paper does not slow the decline. It forecloses on the only audience that could have reversed it. That is the trade. It is the trade most of the industry has already made. It is still being made, in city after city, by people who mistake a subscription list for a plan.