By Dan Kennedy • The press, politics, technology, culture and other passions

Medill, after prodding, begins updating and correcting its map of news deserts

The Medill map of news deserts from November 2023

If there is an ur-source of data about the extent of the local news crisis, it is surely the map of news deserts compiled by journalism researcher Penelope Abernathy. First at the University of North Carolina and more recently at Northwestern University’s Medill School, Abernathy has produced perhaps the single most influential work on the local news crisis.

Ellen Clegg and I have cited her topline numbers — 2,900 newspapers, mostly weeklies, closed since 2005; 43,000 journalism jobs lost over the same period — countless times in talking about our book, “What Works on Community News.” Abernathy is also a professional friend who was kind enough to blurb our book and who’s appeared on our “What Works” podcast.

Compiling the sort of data needed to produce a reliable presentation like the one put together by Abernathy and her colleagues, though, is incredibly hard work. A lot of it depends on the methodology you use. New projects constantly come online; others flicker out. As such, I’ve tended to regard her study as most useful when viewed from 40,000 feet, less useful when you take a look at county-by-county results.

The most recent Medill data is from November 2023. Alice Dreger, who writes the blog Local News Blues, has been pointing out flaws in the numbers all year. Among other things: The Baltimore Sun, a legacy for-profit newspaper, and The Baltimore Banner, a digital nonprofit, were originally excluded; East Lansing Info, a sizable nonprofit in Michigan that Dreger helped found, was not on the list; and several projects serving Black communities, such as  AfroLA and Baltimore Witness, were missing.

I’ve noticed problems as well, especially with startups that were founded within the last couple of years in response to Gannett’s decimation of its weekly newspapers in Greater Boston. Some may have popped up after Medill’s deadline.

Now, Dreger reports, Medill is starting to make corrections after initially being unresponsive. That’s good. In fact, I would suggest that the Medill report and others like it should always be regarded as works in progress, subject to additions, deletions and other edits. I maintain a database of independent local news organizations and news-oriented public access stations in Massachusetts, and I’m always making changes to it.

I don’t think the problems Dreger found detract from the overall value of the Medill database. I do think it’s important to regard it as a living, breathing representation of the local news situation across the United States, valid in its essentials but always subject to change and updates. And kudos to Dreger for getting the ball rolling.

Leave a comment | Read comments

Discover more from Media Nation

Subscribe to get the latest posts sent to your email.


Dan Kennedy of McSweeney’s checks in on the rest of us


Margaret Sullivan on The Washington Post and the demise of the public editor


  1. “I don’t think the problems Dreger found detract from the overall value of the Medill database.”

    I’m afraid I have to disagree with you on this one, Dan. I think the many inaccuracies coupled with the seeming lack of interest to correct them are a big cause for concern.

    On the data: No one in a newspaper would produce such sloppy work. If the NY Times did something that slipshod, there would be a huge controversy. In fact, it would be the kind of scenario you routinely critique here 🙂

    On the lack of corrections: If you make mistakes, fess up and correct them. There’s no excuse for letting this mess fester.

    On the overall value: Being mischaracterized or omitted is a big problem. A lot of funders and potential advertisers will draw from this database because it comes from such an “authoritative” source.

    You say you have drawn on Medill’s headline numbers for your own work. Can you be sure those are valid? I’d be wary.

    I agree that a nationwide project like this is an ambitious undertaking. It would be of value if it were accurate, or at least updated. But a drive-by survey like this is not the answer.

  2. I agree with Mr. Muldoon. Being able to spot numerous individual errors calls into question the credibility of the entire map. (Who says, “Well, yes, I know the Baltimore Sun is missing but it’s an excellent map”?) I remember when Michelle Ferrier compiled a news desert map even before Penny Abernathy did. It included, for example, dots representing new dailies that had been started in the northwestern corner of Kansas. Yet anyone who knows anything about northwest Kansas, or could quickly find out basic information about the northwestern corner of Kansas, knows that the chances of any new daily being started there were, um, less than zero. (It’s much more likely that dailies had been shut down there!) I just laughed. If you’re going to do a national map that purports to show every news organization, you have to put in the time and effort to make it as complete and accurate as possible, and constantly update it. You need to get into the weeds. Then when you say the USA has exactly 252 (or whatever) news deserts (counties with no local news media), people believe the number is really 252.

  3. angelia herrin

    Dan, Penny retired from Northwestern last year. She is still active in research but it would be unfair to lay the problems on the current data base on her door step.

    • Dan Kennedy

      Angelia, Penny only retired in May, and she was in touch with me last year for some information to include in the news-desert map. In any case, it was obviously a group effort. I just hope they keep correcting it, because there really are a lot of problems with it.

Powered by WordPress & Theme by Anders Norén