Methods Appendix

How We Created the Database of Oregon News Outlets

As described in the main report, the data on Oregon’s news outlets presented in this report are based on a data set consisting of 241 media outlets, including newspapers, magazines, online news sites, radio stations, and television stations that we concluded regularly produce original local news content (further defined below). It represents our best efforts to catalog all Oregon news outlets meeting that definition at the time this report was written (mid-2021 – mid-2022).

We cannot say with certainty that this is a complete census, since the media landscape is constantly changing and some outlets that publish news infrequently, do not have an online presence, or do not reach a broad local audience may not be picked up by the methods we used to create this media map. We welcome feedback here as to how our database could be improved.

Defining the Scope. Although some communities in Oregon are served by news media outlets from beyond the state’s borders, we decided to limit our ecosystem assessment to news outlets based in Oregon. While similar studies in a few other states have addressed how media in their neighboring states may offer relevant content to state residents, those states are more densely populated than Oregon and their media markets may overlap more than in Oregon. There are certainly some media outlets in Washington, Idaho, and California that serve some areas of Oregon as well, but overall these appear to be limited. 

As a note, there are two possible methods for identifying relevant news outlets in a given state, based on either physical location or on locales reached (via broadcast signal, circulation, etc.). Both of these options take a top-down approach to media mapping. A bottom-up approach that maps media ecosystems based on where audiences in a state actually go to get various kinds of news has yet to be developed (Stonbely, 2021), particularly at a scale like the one used in this study. 

The Initial Database. Following the method used in another state-level media mapping project,1Philip M. Napoli, Matthew Weber, Katie McCollough, and Qun Wang, (2018), “Assessing local journalism: News deserts, journalism divides, and the determinants of the robustness of local news.” we began mapping Oregon’s media landscape by consulting the Cision media database. Cision, a platform that catalogs media outlets, programs, journalists, and influencers, offers a search function that can retrieve all media outlets registered with their site by location. This search produced a dataset of 1,250 media entities based in Oregon. We then culled that list by removing specific programs offered by a media outlet (e.g., Tillamook Today – KTIL-FM), individual people (e.g., freelance journalists), duplicates (e.g., “1859 – Oregon’s Magazine” and “1859 – Oregon’s Magazine Online”), and defunct outlets (e.g. shuttered newspapers or inactive blogs).

Importantly, Cision has coverage gaps, and other researchers have therefore recommended the use of supplementary databases in addition to Cision2Philip M. Napoli, Matthew Weber, Katie McCollough, and Qun Wang, (2018), “Assessing local journalism: News deserts, journalism divides, and the determinants of the robustness of local news.”. Accordingly, we also compiled data from the US News Deserts database, Unite OR, Oregon Newspaper Publishers Association, Project Oasis/Local Independent Online News Publishers (LION), Institute for Nonprofit News, American Journalism Project, and Poynter, as well as adding outlets we found through convenience sampling (e.g. the researcher hears of or knows of an outlet that has yet to be added) and snowball sampling (e.g. collecting data on one outlet leads the researcher to another).3We also compared our dataset with the Metric Media dataset that identifies right-wing AI journalism pages. None of these sites were included, as they do not meet the criteria for original, local civic news content as outlined in our coding process.. All said, our combined database from the eight existing databases and two additional forms of sampling resulted in 645 individual media outlets.4 Additional radio stations were later added to the dataset as we accounted for the various rebroadcasts and member stations of KLCC, Jefferson Public Radio, and Oregon Public Broadcasting stations throughout the state. Data collection, from identifying the first sources to completing the coding process, was conducted from July 2021 to August 2022. 

Variables and Coding Process. The following pieces of information were collected for each media outlet: 

  • Outlet name and website link. 
  • Address (as precise as possible; e.g., street name or crossroads, city, zip code). 
  • Medium (i.e., blog/online, circular, magazine, media & community organization, newsletter, newspaper, radio, television). 
  • Frequency of distribution (i.e., 2x/month, 2x/week, 2x/year, annually, daily, every other month, every other week, irregular, monthly, quarterly, semimonthly, semiweekly, triannually, triweekly, weekly). 
  • Keywords related to the outlet (e.g., business, politics, Spanish-language, Indigenous). 
  • Chain/owner of the media outlet. 
  • Year the outlet was founded. 
  • Circulation/audience size for print publications. 
  • Contact information (name, email address, title, phone number). 
  • Source of the information (e.g., Cision, LION). 

This information was extracted from the database in which we located the outlet (e.g., Cision, US News Deserts) and/or the outlet’s own website. If the information was unavailable through either method, it is listed as missing in our database.  

Winnowing for Relevant Outlets. In order to identify which of these 645 outlets regularly produce local news, two independent coders examined each of the media outlets’ websites to identify recently published content that met the criteria set forth below. Both coders had to agree on inclusion, with borderline cases discussed with and verified by a third researcher. Coding was conducted from December 2021 to May 2022 and resulted in 241 retained media outlets, represented by the interactive map of media outlets in the main report. 

Some platform-specific decisions needed to be made. For newspapers and magazines, we included monthly and bimonthly publications, business journals, lifestyle magazines and zoned editions, among other publication types, so long as they met the criteria for producing regular, local, original civic news reporting described below. In this we differed from the approach taken by the widely-cited News Deserts report, which excluded some of these publication types.5 Abernathy, P. M. (2020). “News deserts and ghost newspapers: Will local news survive?”University of North Carolina Press,  p. 114. The author states, “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.” In the case of local newspaper chains, we allowed that the content produced by one zoned edition or outlet in a chain counted as local, original content for other outlets and editions in the chain–consistent with our overall definition of “local” news as news pertaining to Oregon.

For broadcast outlets, as described in the main report, commercial radio stations were removed at this juncture, based on national studies showing generally low levels of news content aired by such outlets.6Abernathy, “News deserts…”, p. 67; Jessica Mahone, Wun Wang, Philip Napoli, Matthew Weber, & Katie McCullough, “Who’s producing local journalism? Assessing journalistic output across different outlet types” (August 2019), DeWitt Wallace Center for Media & Democracy, p. 3. Public radio stations and television stations were examined in a similar manner as other forms of content, by visiting the station’s website and checking for recent, original, local news content. Radio stations and television stations were considered “daily” content producers, and as with newspaper chains, stories written by the same owner (e.g., KVAL and KMTR) counted as original for members of the chain. 

The Final Data Set: Regular Production of Original, Local, Civically-Relevant News

Our final database, upon which all data presented in the main report are based, includes Oregon-based outlets that 1) publish at least some local news content at least monthly, and b) produce original local journalism c) covering issues of local (i.e. Oregon or local Oregon communities) civic relevance. Importantly, due to resource limitations, only outlets whose content could be accessed online are included in this database.

“Regularly produced” news. We defined “regularly” producing content as7These standards were established after a pilot review of 40 sites. We wanted to include as many news and information sources as possible, and purposefully developed a low publishing frequency standard for this reason.

  • Dailies and weeklies that publish an average of at least one local original story on issues of civic relevance each week. A maximum of two weeks’ worth of content was reviewed by coders to determine whether to include the outlet. 
  • Bi-weeklies that publish an average of at least two local original stories on issues of civic relevance per month. A maximum of one month’s worth of content was reviewed by coders to determine whether to include the outlet.  
  • Monthlies that publish an average of one local original story per month. A maximum of two months’ worth of content was reviewed by coders to determine whether to include the outlet.8Not all types of publications were included in this assessment. While we are aware that local community organizations, cultural organizations, government agencies, nonprofits, businesses, industry groups and other organizations and people produce newsletters, social media posts, websites and other types of valuable content, we do not have the staff to locate and evaluate all of these types of content sources. For this reason, the focus of our review is news outlets with a website.

This is a broader and more lenient standard than that taken in a widely-cited Pew Research Center study, in which “All providers identified as producing at least weekly content were analyzed” (Pew, 2015, p. 10). 

“Original local content.” Content was counted as “original local content” if it was a “reported” story, meaning the content creator gathered information from spokespersons, experts, documents and other sources and attributed information in the story to those sources. If an outlet appeared to only run stories produced by other organizations, including wire stories and press releases, we did not include that outlet in the database.9In cases where content appeared to be a press release but was not labeled as such (which was not infrequent in smaller publications), we performed Google searches to see if the content source could be identified. We did include content that was produced by another outlet in the same chain or owned by the same company. Similarly, if an outlet appeared to only publish opinion pieces, essays and editorials, press releases, or algorithmically-generated content, we did not include it.10Often these were labeled as opinion pieces, essays or editorials, but sometimes they were not. In cases where the author expressed opinion, or recommended action to readers, politicians or others, or declared developments or issues to be good or bad, these pieces were considered essay, opinion or editorial. We also did not count marketing or sponsored content, or conflict of interest content (for example, someone writing about themselves or their organization, or a business writing about their business or industry) as original local content. 

“Civically relevant news.” A broad range of topics is included in our definition of civically relevant content, including coverage of local government and politics, philanthropy and volunteering, environment, education, health, fire, emergencies and weather, crime, courts and law, economic issues, transportation, development and real estate, housing, diversity and inclusion, poverty and inequities. Also included are local sports, local arts, and local entertainment stories––all offer opportunities for residents to gather, get to know one another and create community. Profiles of local businesses, artists, coaches and other newsmakers in the community are included. Not included are consumption-oriented articles, including for example articles that suggest wines or restaurants you should try, places you should travel and items for consumers to try.

Our coding choices were informed by the work of other research teams who have mapped local news ecosystems by looking for coverage that “is original, is about the local community, and addresses critical information needs.”11Napoli, P. M., Stonbely, S., McCollough, K., & Renninger, B. (2017). Local Journalism and the Information Needs of Local Communities: Toward a scalable assessment approach. Journalism Practice, 11(4), 373-395.  p. 373. We define civically relevant content (the terminology we use here, instead of “critical information needs”) as content that helps residents learn about and participate in their community, work together to address shared problems, “make wise decisions that will affect the quality of their lives”12Abernathy, “News deserts,” p. 18. and “live safe and healthy lives.” 13Friedland, L., Napoli, P., Ognyanova, K., Weil, C., & Wilson III, E. J. (2012). Review of the literature regarding critical information needs of the American public, p. v. choices were also informed by other news mapping projects, such as one by the Pew Research Center, which included as “civic” topics: “government, police/fire departments, campaigns/politics, development, court/legal system, business/economics, education, transportation, health/medicine, environment, charity/philanthropy, drug policy, … religion, and science/tech.”14Pew Research Center (March, 2015), “Local News in a Digital Age”. A similar project by an experienced research team at the University of Wisconsin (Friedland et al. 2012, p. v), used the following categories of information, most of which we included:   

1. emergencies and risks, both immediate and long term; 2. health and welfare, including specifically local health information as well as group specific health information where it exists; 3. education, including the quality of local schools and choices available to parents; 4. transportation, including available alternatives, costs, and schedules; 5. economic opportunities, including job information, job training, and small business assistance; 6. the environment, including air and water quality and access to recreation; 7. civic information, including the availability of civic institutions and opportunities to associate with others; 8. political information, including information about candidates at all relevant levels of local governance, and about relevant public policy initiatives affecting communities and neighborhoods.

We chose not to include content that was in the form of simple listings, such as bus schedules, job listings, and calendar listings promoting upcoming events. So, content telling audiences about “opportunities to associate with others” (a criterion adopted by another research team15Friedland et al., 2012, p. v.) would be included if it were an original story about an upcoming local event, but would not be included if it were a calendar listing. And while we would count a story about a series of new job training classes, we would not count calendar items that merely list the hours and addresses of job trainings.  

Database management. Datasets were maintained in a private Microsoft Teams server that was only accessed by the authors of the report. As all the data are based on publicly available information and did not involve human subjects, no Institutional Review Board approval was required for the compiling of the database. Readers may access the database of 241 Oregon news outlets here

Ownership Type. For each media outlet, we used information from either the relevant data source (e.g., Cision), the outlet’s own website (e.g., “About Us” page), or from Internet search queries (e.g., “X news owner”) to determine the owner of the outlet. Once collected, these were re-classified16Owner name is included in our dataset as a separate variable. Using the interactive map, owner name can be viewed for each media outlet entry. into one of the following categories17See Johanna Dunaway and Regina G. Lawrence, “What predicts the game frame? Media ownership, electoral context, and campaign news,” Political Communication 32, no. 1 (2015), pp. 43-60. The ownership categories shown here built on the work of Dunaway and Lawrence, though we added a non-profit category.:

  • Privately-owned single-media holding: A media outlet owned by a private company with a single holding (i.e., asset, company, or business). An example of this model from our dataset is The Nugget Newspaper in Sisters, which is owned by The Nugget Newspaper, LLC.
  • Small local chain: A media outlet owned by a private company that has multiple holdings within Oregon. An example of this model from our dataset is any newspaper owned by Pamplin Media.
  • Large geographically diffuse chain: A media outlet owned by a private company that has multiple holdings within and outside of Oregon. An example of this model from our dataset is any newspaper owned by EO Media Group (as they have holdings in Oregon and Washington).
  • Publicly traded and shareholder controlled corporation: A media outlet owned by a corporation in which controlling stocks are not held by a single owner or family. An example of this model from our dataset is any television station owned by Sinclair.
  • Non-profit: A media outlet that is supported by non-commercial funding (e.g., donors, federal funding), is a registered 501c organization, or is otherwise not affiliated with a corporate model of ownership. There are 10 sub-types of non-profit outlets in the dataset. These are: Employee-owned (i.e., Burns Times-Herald), environmental/science (e.g., Columbia Insight), Jefferson Public Radio affiliates, KLCC affiliates, Oregon Public Broadcasting affiliates, non-specific/nonpartisan (e.g., KBOO FM), political and grassroots organizations (e.g., KEPW FM), religious (e.g., Catholic Sentinel), tribal/Indigenous (e.g., Smoke Signals), and university/schools (e.g., Student Insurgent).

Mapping the Data. The density heat map provided in the report consists of three layers:

  1. County-level population density: Counties are represented as either frontier, rural, or urban in the first layer of the map. This classification was adapted from one similarly used by Oregon Health & Science University and the Oregon Office of Rural Health.
    1. Frontier: Counties with 6 or fewer people per square mile.
      1. Baker, Gilliam, Grant, Harney, Lake, Malheur, Morrow, Sherman, Wallowa, Wheeler.
    2. Rural: Counties with more than 6 people per square mile but fewer than 100 people per square mile.
      1. Clatsop, Columbia, Coos, Crook, Curry, Deschutes, Douglas, Hood River, Jackson, Jefferson, Josephine, Klamath, Lane, Lincoln, Linn, Tillamook, Umatilla, Union, Wasco.
    3. Urban: Counties with more than 100 people per square mile.
      1. Benton, Clackamas, Marion, Multnomah, Polk, Washington, Yamhill.
  2. Regional dividing lines: Four dividing lines are placed on top of the map to signal where regional geographic distinctions are made. This mapping was adapted from one provided by the Oregon Values and Beliefs Center.
    1. Eastern Oregon: Counties located in the easternmost segment of the state.
      1. Baker, Crook, Gilliam, Grant, Jefferson, Morrow, Sherman, Umatilla, Union, Wallowa, Wasco, Wheeler.
    2. Coastal Oregon: Counties located in the westernmost segment of the state, alongside the Pacific Ocean coastline.
      1. Clatsop, Coos, Curry, Lincoln, Tillamook.
    3. Southern Oregon: Counties located in the southernmost segment of the state.
      1. Douglas, Harney, Jackson, Josephine, Klamath, Lake, Malheur.
    4. Metro/Valley: Counties located in the Willamette Valley and metropolitan areas of the state.
      1. Benton, Clackamas, Columbia, Deschutes, Hood River, Lane, Linn, Marion, Multnomah, Polk, Washington, Yamhill.
  3. City-level heat mapping: News outlets are marked by the city where they’re based, with a gradient indicating how concentrated outlets are in that city/metropolitan area (green indicating less concentrated and red indicating more concentrated).


In addition to mapping Oregon’s news outlets, we wanted to understand their current context, challenges and innovations—and to understand how well Oregon’s media are serving Oregon’s many communities. So we interviewed a broad swath of journalists, community organizations and civic leaders, asking them these questions, among others: 

  • Where are the big gaps in providing news and information to Oregon’s communities? 
  • What are the new, emerging sources of news and information – either as competitors, concerns, or as exciting developments? 
  • How do you see trends playing out in the mid- to long-term? What will the local media landscape in Oregon look like 20 years from now? 

We spoke on the record with 28 sources in Portland and around Oregon. Seventeen of our sources were content producers, newsroom managers, publishers of local news startups and others working in media and media support organizations. Of the others, one works in local government, four are researchers, and seven work with foundations, community services organizations or advocacy organizations.  

Interviews were conducted via Zoom. Subjects gave us permission to record the interviews, and were told that we would share with them any direct quotations we’d like to use in the report before publishing it. 

As a validity check for this research, we also did some limited “respondent validation” in which we asked some key sources to review sections of the report or the entire report, and to comment on whether the findings resonate with them, whether the report contains gaps, and to offer follow-up observations as needed.18Bock, M. A. (2011). Newspaper journalism and video: Motion, sound, and new narratives. New media & society, 14(4), 600-616. 



We thank all the people who shared their valuable time to speak with us about the state of local news, information, and civic life in Oregon. We thank the following individuals for speaking on the record:

  • Therese Bottomly, Editor and Vice President of Content, The Oregonian/Oregon Live
  • Ben Bowman, co-founder Oregon 360 Media
  • Adam Davis, Executive Director, Oregon Humanities
  • Adam Davis, co-founder, Oregon Values and Beliefs Center
  • Caitlin Baggott Davis, Executive Director, North Star Civic Foundation
  • Jacob Fenton, Editor, The Portland Record
  • Kevin Frazier, co-founder Oregon 360 Media
  • Jackleen de La Harpe, founding Executive Director, Underscore
  • Lisa Heyamoto, Director of Teaching & Learning, Local Independent Online News Publishers 
  • Susan Hess, Publisher, Columbia Insight
  • Morgan Holm, Chief Content Officer, Oregon Public Broadcasting
  • Jessica Lagunas, Arts & Culture Manager, Latino Network
  • Chelsea Marr, owner, Columbia Gorge News
  • Jonathan Maus, editor and publisher, Bike Portland
  • Daniel McArdle-Jaimes, Public Information Officer, City of Portland
  • Bruce Poinsette, freelance writer, educator, and organizer
  • Mac Prichard, Founder and President, Prichard Communications
  • Greg Retsinas, News Director, KGW
  • Kaia Sand, Executive Director, Street Roots
  • John Schrag, Executive Editor, Pamplin Media Group
  • Lee Shaker, Associate Professor, Portland State University
  • Quinton Smith, owner and manager, YachatsNews
  • Lee van der Voo, journalist and former managing director, InvestigateWest
  • Amaury Vogel, Associate Executive Director, Oregon Values and Beliefs Center
  • Heidi Wright, Chief Operating Officer, EO Media Group
  • Allie Yee, Co-Executive Director, APANO Communities United Fund
  • Les Zaitz, editor and publisher, Malheur Enterprise
  • Mark Zusman, editor and publisher, Willamette Week