Taking a closer look at the numbers behind a major new study of the local news crisis

For the past 16 years I’ve been reporting on the decline of local news and on efforts to offset it. But though it’s simple enough to spout anecdotes, it can be more challenging to come up with hard numbers, though some have tried.

Sign up for free email delivery of Media Nation — and become a supporter for just $6 a month.

The latest attempt dropped last week: a comprehensive study by Rebuild Local News and Muck Rack, the latter a platform that connects journalists and public relations professionals. I’ve been looking over some of the findings this week, and what’s interesting is that it’s based entirely on data from millions of articles published during the first three months of this year. That means it’s not dependent on the vagaries of counting news outlets by hand, but it also means the researchers had to pile assumption upon assumption and then hope they got it right. I think they did for the most part.

Corey Hutchins sums up the findings for Nieman Lab. According to the topline results, there are 75% fewer journalists working in local news today than there were in 2007, when the Columbia Journalism Review published a story about a similar effort. So let’s pause right there and acknowledge that the 75% figure is based on that earlier study being valid. In any case, the Rebuild Local News/Muck Rack effort, titled “Local Journalist Index 2025,” counts 26,840 “local journalist equivalents,” or LJEs, working across the U.S. That amounts to 8.2 LJEs per 100,000 population, down from 40 in the earlier study.

“This is not a measure of people doing regular accountability reporting,” Steven Waldman, the director of Rebuild Local News, told me by email. “It includes sports, opinions, local events — anything that someone got a byline to cover. A key next step is for this data to be used in a deeper study (probably using AI) to do content analysis.”

Refining the data

The new study is accompanied by a long section on methodology. Among other things, we learn that non-bylined stories are not counted, print newspapers without websites aren’t counted, stories on local television news that aren’t repurposed for digital aren’t counted, and freelancers are counted as 0.1 of an LJE. Attempts are made to filter out aggregators as well. Then there’s this, which the researchers refer to as the “metro adjustment”:

The formula takes the number of Local Journalist Equivalents associated with the largest city in a metro area and assigns 60.16% of that number to the county containing that core city. Then, 26.44% is assigned to the suburbs of that core city. The remaining 13.4% is then divided among all counties in the state except the county containing the core city. Suburbs get a sliver of that number depending on the other factors in that state, so they sometimes end up with a hair more than 26.44%.

In other words, for those of us who live in the Boston area, reporters who work for The Boston Globe, the Boston Herald and, presumably, GBH News, WBUR and local television stations are partly counted as covering the suburbs and, to a lesser extent, the rest of the state as well. The specificity of that formula is based on a study of the Denver news ecosystem, where the core city, the suburbs and the state as a whole are served by The Denver Post, The Colorado Sun, Colorado Public Radio and a variety of broadcast outlets. Having reported on Colorado for our book, “What Works in Community News,” I agree that those city-based news organizations also provide some valuable coverage of the suburbs and the state. I cannot attest to whether 39.84% is the right percentage to use for coverage outside the city. I mean, are they sure it’s not 39.83%?

Kidding aside, I think skepticism about any study is warranted, and I’ll cite a few examples. The oft-updated news desert study by the Medill School at Northwestern University, which relies on a hand count, is often held up as the gold standard, and it is very good. My co-author, Ellen Clegg, and I often refer to its findings that more than 3,200 newspapers have closed since 2005. But it’s not perfect, and at times Medill has had to make corrections in response to critics.

Statistical assumptions can produce some curious results as well. In our book, I wrote about a study of the news ecosystem in New Jersey that found, among other things, that the community of Montclair was served by 42 news outlets, 19 of which were “local news originators.” But once you weeded out the statewide Star Ledger of Newark and broadcast outlets, which weren’t providing basic coverage of Montclair, you were left with four or five outlets at most, which shows that the question you ask is at least as important as the answer. (And yes, I realize that even four or five is unimaginably robust compared to what most places have.)

In examining the assumptions made in “Local Journalist Index 2025,” I took a special look at Middlesex County in Massachusetts, where I live. The data showed 192.7 LJEs, or 11.9 per 100,000. That may not sound like much, but it nevertheless struck me as high. Our What Works project on local news maintains a database of independent outlets in Massachusetts, and there are only a handful from Middlesex County, most employing one, two or three journalists. The chain-owned papers in the county are inconsequential. It seems to me that some of that must be the “metro adjustment” from Boston media that I’ve mentioned, which would certainly pertain to the inner suburbs of Cambridge, Arlington, Somerville and others. Plus Newton, of course, which has always gotten disproportionate attention from the Globe. But Shirley? Tyngsborough? Groton? Not so much.

When I asked Steve Waldman about the numbers for Middlesex County, he suggested I look at “work-adjusted journalists” per 100,000 rather than “metro-adjusted journalists,” which presumably would eliminate the Boston-media effect. That brought down the number from 11.9 to 8.7, which seems a bit more reasonable. “By the way, one of the things we want to work on is refining that metro adjustment,” he said. “Did we make it too high? Too low? Is it the right concept?”

Counting counties

Which brings me to another issue I have with the study. The authors note that county government in Connecticut gave them a problem, but it barely exists in Massachusetts. If most municipal functions are run at the county level, then a handful of reporters for a county might be able to provide real accountability journalism. In Massachusetts, though, we need on-the-ground coverage in every one of our 351 cities and towns. A similar situation exists in New Jersey, which has 564 municipalities and little in the way of county government.

The data were able to capture the news wars taking place on Martha’s Vineyard (Dukes County, 93 LJEs per 100,000) and Nantucket (Nantucket County, 36.1), each of which is served by two news outlets — two weekly papers in the Vineyard’s case. But I’m puzzled by Berkshire County (38.5 LJEs), which is covered by The Berkshire Eagle, a fine daily paper, and some much smaller independents. That struck me as high.

Another surprise was New Haven County in Connecticut, which has a total of 54.5 LJEs, or just 6.3 LJEs per 100,000. The county is home to the New Haven Independent, which I’ve been following for years and which I regard as one of the very best local news organizations in the country. On the other hand, the Independent has always done a lot with just four or five full-time staff members (plus a number of paid freelancers), and it’s tightly focused on the city of New Haven. The city’s population of 135,000 is less than 16% of the county’s 864,000. The Independent also has a sister site northwest of New Haven called the Valley Sentinel Independent, but that is minimally staffed. So I guess the 6.3 figure makes sense, but it fails to account for how well the city of New Haven is covered.

Waldman emphasized that there are limits to such a report. “This study really can’t get at the quality or nature of the coverage,” he said. “It’s just designed to give a sense of scale — over time, across the country, and among different counties.” Indeed, one of my frustrations with our What Works database is that there is no way of singling out the handful of outlets that are producing truly outstanding local coverage.

Nevertheless, “Local Journalist Index 2025” is a valuable addition to the body of work documenting the local news crisis — and it will become only more valuable in the years ahead, because then we’ll be able to make comparisons over time.