Methodology

The findings in this report are based on information in a comprehensive proprietary database that is created and maintained by the Center for Innovation and Sustainability in Local Media at the University of North Carolina at Chapel Hill. Over the past six years, faculty and researchers in the Hussman School of Journalism and Media collected data on more than 9,000 local newspapers, 1,400 public broadcasting outlets, 950 ethnic media and 525 digital sites. The information is derived from a variety of industry and government sources, supplemented with extensive research and reporting, fact-checking and multiple layers of verification. UNC maintains six separate databases on newspapers that were published in 2004, 2014, 2016, 2017, 2018 and 2019, plus two databases on local digital sites published in 2018 and 2020. With the 2020 report, we are also creating two new databases with information on ethnic media and public broadcasting outlets.

Our first two databases (of newspapers published in 2004 and 2014) were created in 2014, as an effort to track the fate of the country’s local newspapers over the previous decade, a tumultuous period in the industry. Extensive research has concluded that local newspapers have historically been the prime source of news and information that guides decision-making of residents and government officials at the grassroots level on issues such as education, elections and the environment (to name but a few). So, we focused first on finding data on the thousands of small dailies and weeklies in this country, since there was no comprehensive and reliable source of information on them. Our 2016 report tracked the loss of local newspapers. Our 2018 and 2020 reports added a new layer of research as we sought to assess the health of the news ecosystem in a community. In 2018, we added information on local digital sites, and with this 2020 report, we are compiling our first list of ethnic media and public broadcasting outlets.

  • Newspapers: Information on individual newspapers in the database has been cross-referenced with at least four sources. Our database was initially created using statistics gleaned from two industry databases: Editor & Publisher DataBook (published 2004-2017) and E&P data accessed online for the years 2016-2019, as well as proprietary information collected and provided by the consulting firm BIA/Kelsey for the years 2004 and 2014. Researchers then verified these data with information obtained from 55 state, regional and national press associations, and our own extensive independent online research, ascertaining the content of newspapers by checking websites and print versions, and also interviewing editors and publishers when appropriate.
  • Local News Sites: Our list of local news sites was compiled by merging lists provided by the Local Independent Online News (LION) organization, the Institute for Nonprofit News (INN) and the Google-sponsored Project Oasis. As with newspapers, researchers then visited all of the websites to verify that the content was updated regularly and that the domain was still active.
  • Ethnic news outlets: Our list of ethnic media outlets in the United States was compiled and cross-referenced, using lists of ethnic newspapers in the Editor & Publisher DataBook (2004 and 2018), the Craig Newmark School of Journalism’s (CUNY) State of Latino News Media, the National Newspaper Publishers Association (NNPA, an association of African American newspapers) and the National Association of Black Owned Broadcasters (NABOB).
  • Public broadcasting outlets: The list of public radio and television stations came from data supplied by Public Broadcasting Service (PBS), National Public Radio (NPR), American Public Media (APM) and Pacifica. To determine if a station produced original content, we reviewed individual station’s schedules and contacted individual stations with questions.
  • In addition, we used media information in the report contained in numerous surveys and reports by, among others, the Knight Commission on Trust in Media and Democracy, the Gallup organization, and the Pew Research Center.

Layers of demographic, political and economic data from government sources were also added to the database. By visiting this website – usnewsdeserts.com – and using the interactive maps, researchers, as well as interested citizens, can drill down to the county level in all 50 states and compare how communities across the country have been affected by the closing of local newspapers.

Despite adding multiple layers of verification, we realize the UNC Database is still prone to errors inherent in any large database, particularly one that depends in part on surveys and the accurate feedback of respondents. When we spot errors, we correct them in the database and will continue to update our analysis as new information becomes available. If you detect an error, please fill in and submit the corrections form available on our website. In general, we attempt to update our most current year’s database on a quarterly basis.

Since these are some of the most comprehensive and up-to-date databases on news organizations, we make every effort to share them with serious academic and industry researchers who are pursuing related or relevant topics. If you would like access to our databases of news organizations, please click here to fill out a request form.

Building and Refining the Newspaper Database

By surveying all 50 state press associations in 2017 and 2018, we identified 300 local newspapers that existed in 2004 but were not listed in national industry databases. Therefore, in our 2018 report, we adjusted upward the number of newspapers in our 2004 database – to 8,891 newspapers (1,472 daily; 7,419 weekly). Our 2020 report keys off the revised numbers for 2004 in our 2018 report. Our 2020 report identifies 6,734 local newspapers in the country – 1,260 dailies and 5,474 weeklies – that were still being published at the end of 2019. Each newspaper in the database has the following variables: year, name, frequency (daily/weekly), number of days published per week, city, state, parent media company and total circulation.

Our research is concerned with identifying local newspapers that provide public-service journalism. Therefore, in addition to using industry and press association lists, we add a fourth layer of verification. We consult both online and print editions of newspapers, analyzing the content of several editions of a paper, asking the following questions. Does the paper, for example, cover local government meetings, such as school boards and county commissioner meetings? Does the paper provide coverage on any of the eight topics identified by the Federal Communications Commission as being “critical information needs”?

Intentionally excluded from our proprietary database are shoppers, community newsletters (which focus on people and events, instead of news), specialty publications (such as business journals and lifestyle magazines), monthly and bimonthly publications, advertising inserts, TMCs (Total Market Coverage publications), and some zoned editions that feature minimal local journalism relevant to the county where the zoned edition circulates. This report assesses the quantity, but not the quality of news generated by local news outlets, a step that would require in-depth analysis of the content. We recommend this as an additional research step to anyone seeking to determine the health of the local news ecosystem in a specific region.

Many states and municipalities have different thresholds for determining if a newspaper is a “paper of record” and therefore eligible to carry legal advertisements. Often that threshold is based on circulation and distribution. We recognize that the income from legal advertising is very critical to small dailies and weeklies. Therefore, we can work with the general counsel at individual press associations to provide qualifying text on state maps that explain the difference in our methodology (which is focused on news coverage) versus the threshold used by government officials to determine if a publication is eligible to receive legal advertising. The counsel for state press associations should contact us with requests for qualifying text concerning the status of publications that meet the threshold for legal advertising.

Defining “News Deserts” and “Ghost Newspapers”

As a result of our extra layers of verification, we expanded our definition of “news deserts” and introduced the concept of “ghost newspapers” in our 2018 report. Previously, we defined a news desert as a community without a newspaper. As a result of the dramatic shrinkage in the number of local news outlets in recent years – without alternative news media to replace them – we have expanded our definition of news deserts to include communities where residents are facing significantly diminished access to the sort of important local news and information that feeds grassroots democracy.

We also noted that a number of traditional stand-alone newspapers had become shells, or “ghosts,” of their former selves. They are no longer providing residents with the news they need to make informed decisions about a range of important issues that could affect their quality of life. We identified two types of ghost newspapers: the once-iconic weeklies that merged with larger dailies and evolved into shoppers or specialty publications, and the metro and regional state papers that have dramatically scaled back their newsroom staffing, as well as their government coverage of inner-city, suburban and rural communities.

UNC removed from the 2018 database almost 600 weeklies that were stand-alone traditional newspapers in 2004, but by 2018 had evolved into shoppers, advertising supplements or specialty publications such as lifestyle magazines or business journals. We did not remove the larger metro and regional state papers, but estimated the number – at least 1,000 – by comparing industry statistics on newsroom staffing and circulation to news articles about the size of an individual paper’s newsroom staffing in 2004 compared with 2018. To determine definitively whether a large daily is fulfilling its civic journalism role of informing a community on important issues, much more research – including in-depth analysis of published content – is needed. Having raised the issue, we leave that to other researchers to determine whether an individual paper is a “ghost.”

As was the case with the 2016 report, because our focus is on local newspapers, UNC also excluded from the 2018 and 2020 reports data on the country’s largest national papers – The New York Times, The Wall Street Journal and USA Today – as well as shoppers, magazines and other specialty publications. In total, 2,196 papers that were closed, merged or morphed into shoppers or specialty publications over the past 15 years were removed from the UNC database to arrive at the 2019 number. Nevertheless, the 2020 tally of 6,734 papers may overstate the number of stand-alone newspapers. Based on UNC’s analysis of papers owned by the largest 25 chains, an estimated 10 to 15 percent of newspapers still listed in industry databases may be, in fact, zoned editions of larger papers. Nevertheless, these zoned editions were not removed from the 2018 tally, since they are still providing news coverage of important events and issues in their communities.

Dealing with Circulation Limitations

There is currently no widely accepted and easily accessible tracking system of online readership data, especially for the thousands of local papers in small and mid-sized markets. Therefore, print circulation is used as a proxy for measuring the decline in both the reach and influence of traditional newspapers.

The print circulation figures in our database come with limitations. Some circulation figures are audited and verified; others are self-reported. Therefore, in our 2018 database, we’ve added additional verification steps and information in an attempt to be as transparent as possible about where we are getting the numbers. We also noted whether the reported circulation is free or paid circulation.

When possible, we use circulation numbers from the Alliance for Audited Media (AAM). AAM is the industry leader in media verification and specializes in verifying circulation metrics for publishers. However, only 13 percent of papers in the UNC Database subscribe to AAM audits. Additionally, the reported AAM numbers for the large dailies often lag behind the audit by a couple of years.

Because news organizations must pay AAM to verify their circulation statistics, many small papers do not use the service and instead self-report. If there are no AAM data, UNC relied on self-reported newspaper circulation from a variety of sources (E&P, state press associations and independent research). Self-reported circulation data are problematic, since UNC researchers observed that a significant number of newspapers report the same circulation across multiple years. However, self-reported numbers are the only option available for many small weekly papers.

U.S. and State Maps

For the 2018 and 2020 report, UNC created interactive maps for the country and all 50 states, plus the District of Columbia. The maps in this report, and in its online version, provide insights into the risk of news deserts in thousands of communities across the country. By visiting http://www.usnewsdeserts.com, researchers can analyze data (demographics, political leanings, number of news outlets) down to the county level for all 50 states. UNC researchers used government data to pinpoint the locations of newspapers as accurately as possible. Often, both the BIA/Kelsey and E&P 2014 databases incorrectly listed the parent company or city location for many newspapers, especially the smaller ones. UNC researchers attempted to review and correct errors. The UNC Database uses the newspaper’s office as the physical address for mapping purposes.

To identify whether newspapers were located in a rural or an urban area, each was assigned to a corresponding group from the U.S. Department of Agriculture’s Rural-Urban Continuum Codes (RUCC) based on the county in which they were located. According to RUCC codes, communities in groups one through three were classified as metro areas. All others were classified as rural. Additionally, U.S. Census information on demographics (income, age, population makeup, etc.) was merged into the databases, as well as information from state election boards and industry sources such as the Local Independent Online News (LION) association. We also overlaid the USDA’s information to locate counties in food deserts. The USDA defines a food desert as “parts of the country vapid of fresh fruit, vegetables, and other healthful whole foods, usually found in impoverished areas due to a lack of grocery stores, farmers’ markets and healthy food providers.”

For national and state maps, click here.

Tracking Sales, Mergers and Closures

UNC tracks changes in a newspaper’s ownership, as well as closures and mergers, through news accounts and press releases. We define a closure as a newspaper that is no longer published and a merger as a newspaper that has been combined with another publication. Often the two merged papers initially have a combined name, but eventually the name of the smaller paper is eliminated. We tracked mergers and acquisitions in the newspaper industry from 2004 to 2020 and assessed corporate strategies by identifying and examining:

  • Publicly available corporate documents, including quarterly and annual reports released by the individual companies and press releases by Dirks, Van Essen, Murray & April and by Cribb, Greene & Cope, two of the leading merger and acquisition firms in the U.S. newspaper industry.
  • Numerous news articles about individual purchases and business decisions.
  • Statements made by executives that were in press releases, news articles or industry presentations.
  • Reports and interviews with industry representatives and analysts.

There are limitations to all of the above sources. Press releases, news articles, statements made by news executives and reports from industry analysts often list by title only the sales of the largest and most prominent newspapers, usually dailies. The weeklies involved in the sale, as well as specialty publications (including shoppers and business journals) are often grouped together and reported as a single number. That is why we try to check all announcements of sales against publicly available documents and corporate websites.

We track updates to the industry through the Twitter account @businessofnews and post important developments on our website here. For the past three years, we’ve attempted to update the current year’s database on a quarterly basis. The final update of the 2019 database occurred in January 2020, when all transactions that occurred in the fourth quarter of the previous year were recorded.

Media Groupings:

Similar to our 2016 and 2018 reports, UNC categorized the largest 25 newspaper owners into one of three categories: private companies, publicly traded companies and investment companies.

  • Private Companies: This group includes large companies, such as Hearst Corp., which own a portfolio of media that include an array of media formats. They not only own print publications, but cable networks and digital enterprises as well. This category can also include smaller companies like Boone Newspapers, which owns fewer than 100 publications in small and mid-sized communities throughout the South.
  • Public Companies: This group includes publicly traded companies such as Gannett, Lee Enterprises and McClatchy.
  • Investment Companies: This category has arisen in the past decade and has a different ownership philosophy and financial structure from the traditional newspaper owners. Owners in this group can be either private or public, but the key distinctions in investment company ownership are the companies’ business philosophies and financial structures, which differ significantly from those of traditional newspaper chains. Companies were classified in this category if they met at least five of the eight characteristics in the chart below:

In the 2020 report, we note that investment companies – such as Gatehouse, Digital First and BH Media – are merging with the few remaining publicly traded companies, such as Gannett and Lee Enterprises, and forming mega-chains, once again altering the news landscape.

Facebook Data

In March 2019 Facebook shared data with four universities – Duke University, Harvard University, the University of Minnesota and the University of North Carolina at Chapel Hill – on its “Today In“ platform, which provides links to articles from local news organizations. To get an idea of the quality of information in the news articles available to North Carolina residents, the UNC team of four researchers focused on the content of the “Today In” stories for the month of February 2019. Our analysis zeroed in on six major metropolitan areas in North Carolina, as determined by the Office of State Budget and Management. By cross-referencing the six major metropolitan areas in North Carolina and the cities that had the “Today In” app at the time, we created a list of 14 counties and 12 “Today In” cities for the analysis. The researchers coded and categorized “Today In” news articles in these cities as to whether they provided users with public service journalism and information about the eight critical information needs (such as transportation, politics and education), as identified by the Federal Communications Commission. In addition, another researcher followed up by tracking the “Today In” news articles available to residents living in three cities in the Research Triangle area – Raleigh, Durham and Chapel Hill – during one week in September 2019.

About the Industry Databases:

Newspapers and Digital Sites: Editor & Publisher began publishing an annual Newspaper DataBook in 1921. The DataBook has information on more than 25,000 companies and more than 160 data fields. Data are collected through mail and email surveys. BIA/Kelsey, a research and advisory company, focused on local advertising and marketing, began tracking newspaper ownership in 2004. The organization employs a telemarketing team that calls individual newspapers and collects information from employee respondents. Data on local digital sites were compiled from two separate lists maintained by the Local Independent Online News (LION) organization, the Institute for Nonprofit News (INN), and the Google-funded Project Oasis.

Ethnic Media: E&P provided data on three categories of ethnic newspapers: African American, Hispanic and Other. Through additional research, we were able to obtain data on the following ethnic newspapers: African American, Asian American, Latino, Native American and Other (including Polish, Russian and Armenian). This was supplemented with the data published by the City University of New York’s Craig Newmark Graduate School of Journalism on “The State of the Latino News Media,” published in June 2019. This report provided information on broadcast outlets and digital sites, as well newspapers. Additional information on African American media was obtained from the National Newspaper Publishers Association (NNPA) and the National Association of Black Owned Broadcasters (NABOB).

Public Broadcasting Outlets: Information on Public Broadcasting Stations comes from a map of member stations published by PBS on May 2018, referencing it with membership data from American Public Television Stations, a nonprofit organization of public television stations. National Public Radio (NPR) provided a list of members station, which was compared with the NPR’s Station Finder Map. Information on American Public Media (APM) was derived from information on two sites, Southern California Public Radio and Minnesota Public Radio. Pacifica radio has an affiliate network of about 200 stations. However, after speaking with Director Ursula Ruedenberg, we included in our database the only four stations that produce original content: KPFA (Berkeley, California), KPFK (Los Angeles), WPFW (Washington, D.C.) and KPFT (Houston).

Next: Additional Resources