REUSING THE NEWS:Duplicating Local TV Content
Danilo Yanich and Benjamin E. Bagozzi,
University of Delaware
August 2025
INTRODUCTION
Local U.S. television news provides citizens with crucial local information, particularly during periods of heightened political, social, or environmental salience. Yet local television news is under increased pressure to consolidate, which may serve to de-localize local news reporting by duplicating news content delivery across local broadcast stations.
This report examines two fundamental questions. Stripped to their bare essentials, they are:
1. Who controls what?
2. Does that control affect news content?
We constructed three databases to address these questions. The databases span multiple characteristics of local broadcast stations and of the actual news content (i.e., transcript text) that each station aired. Our content database included the news broadcasts of 861 local stations in all 210 television markets in the U.S. that presented news content over a three-month period in the fall of 2019. It is the largest such database in existence. We chose a period before the COVID pandemic so that coverage was not affected by a single overwhelming story.
We employed automated text reuse methods to measure the extent to which local broadcast station pairs duplicated (exact text reuse) each other’s news content. We applied a high threshold for duplication: to be considered duplication, 50% of the broadcast news content (excluding sports, weather and commercials) of a station pair had to be an exact match.
KEY FINDINGS
- Duplication occurred in 84 (39%) of the 210 television markets in the U.S.
- These markets accounted for almost 41 million television households (37%) in the country.
- The duplication overwhelmingly occurred inside the Designated Market Areas (DMAs) (86%) versus outside of them (14%).
- There were 96 duplicating station pairs involving 182 unique stations (some stations had multiple arrangements).
- Most duplicating station pairs were connected through Service Agreements (52%), followed by Duopolies (38%) and Common Ownership (stations in two different markets) (10%).
- Stations connected by service agreements exhibited twice the rates of duplication (51%, on average) as did station pairs connected by ownership (25%, on average).
- Smaller DMAs had higher proportions of duplicating station pairs than did larger markets.
- Just four station groups controlled over half (53%) of the duplicating station pairs.
- Nexstar was the most active controller of duplicating station pairs (22%), followed by Gray (17%).
- News-Press & Gazette (8%) and Sinclair (6%) were the third and fourth most prominent controllers of duplicating station pairs.
Markets with duplicating local television station pairs in the U.S.
Tap or click any area on the map to see detailed information about that market.
DMA Duplication Profile
Map by: David Racca, Policy Scientist, Center for Applied Demography and Survey Research (CADSR), Biden School of Public Policy and Administration, University of Delaware
WHY LOCAL TV NEWS?
Local television news remains a formidable news source in the United States, even in the age of the Internet and social media. Local television stations provide crucial local information to citizens, particularly in times of crisis and in local elections that national outlets cannot accomplish.
The Federal Communications Commission identifies ‘health’ as one the eight areas of information critical to communities (Waldman, 2011). The COVID pandemic brought that into stark relief. It was the most severe health crisis in a hundred years and how the public understood its contours determined how we survived it. In no other national emergency in our lifetime had the response from elected officials taken on such contradictory and divisive tones. To say that messages about COVID-19 were mixed is a gross understatement. Citizens needed to know, in real time, how the virus spread, what preventive steps to take, and what to do if they became infected. They needed to know about the status of the virus in their localities because that is where they lived their lives.
By any metric, these are local stories, and citizens need to find local information. But from where? Newspapers are vanishing. Since 2005, the U.S. has lost 3,200 newspapers; there are 206 counties in the U.S. without a news source and 1,561 counties with only one source (Metzger, 2024). Over half of adults get news online but 57% of them think it is inaccurate (Geiger, 2019). Cable programs consist mainly of discussions among panelists. Further, none of these programs address local places unless they are hotspots like New York City. And, cable reaches, based on the pay TV penetration rate, only 64% of the audience in local markets (Leichtman Research Group, 2023).
When COVID struck, the public turned to local television news. This is because it is uniquely able to reach an entire community. Viewership among the 25-54 age group across the top 25 markets in the country rose on average by 31% between February and March 2020; in San Francisco it rose by 38%. Remarkably, viewership increased by 20% for the under-17 age group (Nielsen, 2020a). Neither the networks, cable, nor the Internet provided the public with what it needed to know. In fact, citizens who tried those sources found them woefully inadequate.
The Los Angeles fires in January 2025 spiked local television viewership as residents searched for critical information. Local network affiliates of ABC, CBS, NBC, and Telemundo saw their audience increase between 200 and 300 percent (Eck, 2025). The unscalability of local news and local journalism became crystal clear.
The critical information needs of citizens that were revealed by the COVID pandemic serve as an object lesson for the importance of local television news. The profile of the local television system in the United States makes it a formidable news source. Here are some of the elements of the local news system that make it a compelling feature of the local journalism environment.
Markets: The local television system covers the entire country. Nielsen Research identifies 210 television markets in the United States. Every county in the U.S. is part of a television market (a Designated Market Area (DMA)), although there are a very small number of split counties. The DMA is a geographic area in which people can receive the same radio and television stations, and it is the spatial unit that is used to measure audiences, analyze viewership, etc.
On January 1, 2019 (2019 being the year for which we gathered news content) there were over 110 million television households in the United States distributed among the 210 markets (Nielsen, 2019). Nielsen Research ranks markets by the number of television households they contain. The top four markets are New York (with about 7.1 million TV households) followed by Los Angeles, Chicago, and Philadelphia. At the other end of the list are markets 209 (North Platte, NE (12,830 TV HHs))and 210 (Glendive, MN (3,590 TV HHs)), respectively (Nielsen, 2019).
Audience: The audience for local television news has been decreasing over several years. In 2016, the audience for the local newscasts of the affiliate stations of ABC, CBS, NBC and Fox was 17.5 million viewers. In 2019, the timeframe for this research, the audience was 15 million; it decreased to 12.2 million in 2022 (St. Aubin & Naseer, 2023). However, even in a shrinking news market, local television news still commands a sizable audience. By comparison, the top three cable networks (CNN, Fox and MSNBC) have a combined daily average of just 3.87 million viewers (Johnson, 2021).
Ubiquity: There is much local news. In 2020, there were 1,116 stations broadcasting local news (Papper, 2021). The average number of hours of news that stations air has risen steadily from about 4 hours in 2003 to about 6.6 hours in 2022, a 61 percent increase (St. Aubin & Naseer, 2023). Therefore, on any weekday in the United States, the public has access to more than 7,000 hours of local news broadcasts. Americans take advantage of the content, watching, on average, about 2.5 hours of local television news weekly (Nielsen, 2020b).
Almost two-thirds of Americans get their local news from local TV, more than any other news source.
Local TV as News Source: Americans use several sources for local news. Still, almost two-thirds (64%) get local news from TV news stations, more than online forums (52%); radio (52%) or daily newspapers (33%) (Shearer, Matsa, Lipka, Eddy, & Forman-Katz, 2024). However, those data obscure an important reality: the stories that are consumed online are overwhelmingly produced by legacy mass media sources. For example, almost one-quarter of the public who viewed local TV news in 2019 did so online. For daily newspapers, that proportion was even higher, at 43 percent (Pew Research Center, 2019). In 2024, the proportion of online use by local television consumers rose to 38 percent (Shearer et al., 2024).
That pattern is evident in the prominence of newspapers and local television websites in television markets. The size of the market affects the prominence of newspapers and television as the main sources of local news. In the largest 22 markets in the country (excluding New York and Washington, DC, with their national newspapers), local newspaper websites were the most popular in 14 of them (64%); local TV websites led in the remaining 8 markets (36%). Further, that dominance extended to a sample of 37 smaller television markets (between #25 and #150), where local television led in 23 (62%)and local newspapers led in 13 (35%) markets (Wenger & Papper, 2018).
It is important to note that the proportion of Americans who follow news (of any type) is declining. In 2016, almost eight in ten stated that they followed local news closely; that proportion dropped to about two-thirds by 2024. The proportion of Americans following national news experienced the same conditions, decreasing from 77 percent in 2016 to 68 percent in 2024. Despite the decline in attention, the overwhelming majority (85%) believe that local news outlets are important to the well-being of the community (Shearer et al., 2024).
Trust: Local television is trusted more than any other medium. More than three-fourths (76%) of Americans express a great deal to a fair amount of trust in local television news in their community. That is followed by newspapers (73%). Fewer than half (47%) of Americans trust online news (Guess, Nyham, & Reifler, 2018).
Local TV is trusted more than any other medium for news about the community.
That is in stark contrast to the level of trust in social media news sources for political and election news. Looking at both trust and distrust across social media sources, a significant majority of Facebook users (59%) distrust the medium, while only 12 percent express trust, a distrust-to-trust ratio of 5:1. Almost half (48%) of Twitter users (now “X”) distrusted the site, compared to 12 percent who expressed trust, a ratio of 4:1. For Instagram, the ratio of distrust to trust was 7:1 (42% to 6%) (Hutchinson, 2020).
The Knight Foundation asserts that emotional trust in news organizations is driven by the belief that news organizations care about their communities, report with honest intentions, and are reliable. Using that rubric to assess news organizations, Knight found a contrast between local and national news organizations. Forty-four percent of Americans have a high level of emotional trust in local news organizations, compared with only 21% for their national counterparts. Knight points out that the inverse is also true: less than one-fifth (18%) of Americans have a low level of emotional trust in local news organizations, compared to 41% for national organizations (Knight Foundation & Gallup, 2022).
Civic engagement: Local television is the most prominent news source among local voters and those who are civically engaged in their communities. It is the only media source that surpasses the 50 percent threshold (52%) for citizens who are highly active in the community; social networking sites are at 36 percent, and newspapers are at only 16 percent (Barthel, Holcomb, Mahone, & Mitchell, 2016).
Media affects conversations and 80 percent of respondents reported that local television covers the news of the day that they discuss with family, friends and neighbors (TVB, 2025). Further, even in the competition with digital sources, local TV is the most significant. Of all the digital choices, the conversations about the news of the day are most affected by the digital products (websites/apps) of local television stations (TVB, 2025).
Voters: Voters’ decision-making process involves five steps: awareness (discovering candidates and issues); interest; gathering information; considering whether to vote; and finally, actually casting a vote. In a study of ten competitive states in 2022, in which citizens were surveyed immediately after the polls closed on November 8 through November 19th, television was the dominant source of information consulted at each of those steps, and by a wide margin. Television’s influence was greater than the influence of all other media combined. For example, over half of the respondents (53%) said that TV most influenced their awareness of candidates and issues. Social media was second, with 6 percent. Television hovered around the 50 percent proportion at each stage, with 49 percent of voters saying that television had the most influence on whether they actually cast a vote (TVB, 2023).
TWO FUNDAMENTAL QUESTIONS
Given the importance of local television news in the local media system, we used two fundamental questions to guide our analysis. Both represent immense gaps in the understanding of the local television news broadcast landscape. Stripped to their bare essentials, the questions ask: first, “who controls what?”, and, second, “does control affect news content?”
1. Who controls what?
This question has two parts. The first part focuses on the prevalence of connected stations across U.S. television markets. That connected status can be the result of: (1) a duopoly (two stations in the same market owned by the same station group); (2) a service agreement (stations with different owners have an arrangement in which a brokering station has effective control over a brokered station); or (3) common ownership (two stations in different markets owned by the same station group). What proportion of stations in each market exhibit each connected status?
The second part of the question focuses on control. On its face, the question seems straightforward. Simply examining the ownership of stations, as reflected in the licenses that are granted by the FCC, should certainly indicate “who owns what.” However, ownership does not necessarily mean control. We constructed a station database (described below) to determine the structure of service agreements (SAs) among stations. Service Agreements effectively change the control of stations that are party to them. Although service agreements are disclosed to the FCC, they are not part of the calculus used to determine broadcast ownership policy, because service agreements do not transfer the licenses of the controlled stations to the controlling station.
2. Does control affect news content? If so, how?
In this research, we define “affect” as duplicated content, specifically, the exact matching of the text of broadcast stories. That is, do station pairs that are bound by either ownership or a service agreement air duplicate local news broadcasts? We constructed the content database (described below) to address this question.
LOCAL TELEVISION STATION DATABASE
We developed databases to conduct the analysis. These databases focus on local television stations, markets, and local television news content.
The local television station database contains all 861 stations for which we gathered content. It uses stations as a unit of analysis because the content that we examined was presented on local station broadcasts. Without this database, we would have no understanding of the station characteristics that affected that content.
The database was constructed using several methods. First, we identified the stations in the U.S. that delivered news during the sample period (September 1 through November 30, 2019). That list was derived with the cooperation of Dr. Bob Papper, Distinguished Professor Emeritus at Hofstra University. Professor Papper compiled these data over twenty years of examining the attributes of local television stations; he provided us with the list that he constructed for 2019. In addition to identifying the stations, the database also indicated which of them originated local news and which stations only presented news (the so-called ‘non-originators’). The non-originating stations derived their content from other stations that originated news, through a connected agreement. Three quarters of the stations (n=647) were originators with the remaining one quarter (n=214) being non-originators.
Armed with identification of the stations, we added station characteristics to the database from information purchased from BIA/Kelsey Advisory Services, a media research firm. Those data included: call sign; city of license; state of license; Designated Market Area # (Nielsen ranks television markets from 1 to 210 based on the number of TV households in the DMA); network affiliation; language; owner of the station; service type; total revenue of the station for 2019; total retransmission revenue of the station for 2019; and the percentage of the revenue of the market that each station captured in 2019.
SERVICE AGREEMENTS
Although the BIA/Kelsey data identified the owner of each station, that was insufficient to understand its control profile. Control profile is better understood by examining service agreements among stations. These agreements give some stations control over others without the transfer of a license. It is important to note that because there is no transfer of license in these transactions, service agreements are effectively unregulated by the Federal Communications Commission (FCC). In fact, although service agreements are reported to the FCC, there is no definitive accounting of which stations have agreements with others. Not even the FCC, the entity charged with regulating television station ownership, has a definitive map of that crucial information regarding the control of television stations in the country. However, control information is central to understanding the contour of television station “ownership” profiles.
There are six unique arrangements between broker and brokered local news stations. In any service agreement between stations, the broker station acts as the controlling entity, retaining one of these six forms of management, marketing, or control of news production over the brokered station. These designations are as follows:
- A Joint Sales/Service Agreement (JSA) occurs when a station sells its advertising minutes to a broker station for a fee or a percentage of the advertising revenue. The FCC has recognized that a JSA exists if one station controls 15 percent or more of another station’s advertising time.
- A Local Marketing Agreement (LMA) designates a broker station that controls both programming and advertising of its brokered station.
- A Shared Service Agreement (SSA) occurs when stations combine newsroom tools and services, personnel, and share both facilities and management.
- A Time Brokerage Agreement (TBA) describes the situation in which a broker station broadcasts its programming on its brokered station during specified times.
- A Local News Service Agreement (LNS) designates the content pooling of television news stations between multiple stations within the same market. This content pooling can refer to the sharing of journalists, editors, equipment, and/or content.
- A Resource Sharing Agreement (RSA) is an agreement between two stations to share news footage, scripts, or news-gathering resources.
Each of these types of service agreements represents an arrangement between a broker and brokered station.
There is no complete database on the prevalence of service agreements. We developed that data in the station database to understand the relationship between ownership/control and content. For example, the fact that a station group owns a station’s license does not mean that it controls that station. That is best explained by understanding the service agreements in which the station is involved and whether it is a brokering (controlling) or brokered (controlled) station. We gathered that data from a variety of sources, principally the individual station’s information submitted to the Public Files of the FCC, in addition to the station/station group’s website, trade publications, and media articles. These service agreements were in place in the Fall of 2019, the period for which we secured content data.
It is important to note that we defined the presence of a service agreement narrowly. We did not include service agreements among main stations and low-powered television stations with the same call sign, or digital or satellite agreements. We took this precaution because the inclusion of these arrangements would have overstated the presence of service agreements, which were not the focus of the study.
The information that we developed included the identification of the station with which there was a service agreement, the market in which that station was located, and service agreement type for every service agreement in which each of the 861 stations in the sample participated.
CONTENT DATABASE
For the 861 stations mentioned above, we acquired complete textual transcripts of all news broadcast content from September 1 through November 30, 2019, from TVEyes, a content-gathering company. We chose the Fall of 2019 because it was a period before the onset of the COVID pandemic. During the pandemic, the media functioned in crisis mode. We deliberately examined the performance of the media in a more “normal” environment to avoid overcounting duplication.
This news transcript content was restricted to the following local daily time windows: 0400 to 0700; 1200 to 1300; 1700 to 2000; and 2200 to 2330. We considered all available local news content for the 861 stations during these windows, no matter whether a particular station or program was in English or another language (e.g., Spanish). Most of the content was in English, and our text reuse methods allow us to identify instances of reuse within subsets of transcripts associated with each language.
This news broadcast transcript content was further processed by TvEyes to omit commercials, sports and weather segments, traffic, and non-news programs. Because TvEyes identified and removed these latter content items using automated text-detection methods, their removal was at times imperfect and/or incomplete. Our initial qualitative review of the purchased transcripts suggested that issues of imperfect removal of the above content were minimal, and we removed any remaining non-news content after qualitatively reviewing all retained broadcast titles. This helped to minimize the chances that any remaining ad-based content contributed to our duplication assessments.
TEXT REUSE METHOD
TvEyes provided us with snippets of broadcast news, rather than with complete news broadcasts. Therefore, we combined all station-specific transcript snippets to construct the news broadcasts used in the analysis. That process yielded a total of 419,137 program broadcasts across our 91-day sample. We next applied methods for text reuse detection to each unique program transcript pair to assess the degree of duplication; (1) first within each transcript pair; and then (2) for each local U.S. television station pair. In doing so, we avoided comparing a station’s program transcripts to that same station’s other program transcripts. These methods use automated techniques for identifying plagiarism within text. A determination of text reuse (i.e., duplication) in this context requires a certain proportion of identical textual passages across a pair of transcripts, out of all textual content within that pair of transcripts. This criterion relies upon the order of words in text, rather than word frequencies, and is thus a higher bar for duplicate detection relative to other ‘bag of words’ measures of textual similarity.
Our approach specifically entails the specification of a sequential series of words, represented as “n-grams,” within the broadcast content. We selected an n-gram length that balanced the potential for false positives and false negatives to identify text reuse in news program transcripts. Although there is no single best practice in this respect, studies have historically relied upon trigrams (i.e., 3-grams) for comparable text reuse evaluations (Lyon, Malcolm, & Dickerson, 2001; Lyon, Barrett, & Malcolm, 2006; Flores, Barrón‐Cedeño, Moreno, & Rosso, 2015; Soto et al., 2015). Others have compared the use of n-grams of sizes two, three, and four, reporting negligible improvement beyond n-grams of length four (Morin, Vasek, & Moore, 2021). Less commonly, 5-grams or even 7-grams have been used for text reuse purposes as a means of providing a higher bar for duplicate detection, including in the context of news articles. We selected a relatively high threshold for our analysis---an n-gram of 5 words.
With our primary n-gram lengths assigned, we next extracted every overlapping 5-gram from our program transcripts. These overlapping 5-grams, otherwise known as “shingles,” allowed us to capture all unique sequences of five-grams within our texts (Fig.1). This approach ensured that all potential (5-gram) passages of reuse were accounted for in our assessments. An illustrative example of 5-gram shingles for a single sentence from the corpus of newscasts appears in Figure 1 below. After separately extracting all 5-gram shingles for a pair of program transcripts, all shingles were passed to our text reuse assessment framework. In this step, two transcripts’ pairwise degree of text reuse is represented as the proportion of shingled 5-grams that exhibited an exact match from all 5-gram shingles that were compared for that pair of transcripts, using a Jaccard similarity index. We then chose what proportion threshold of Jaccard similarity constituted text reuse.
Figure 1: Example of 5-gram Shingles for Selected Passage of News Transcript Text
Using the index, we chose a threshold of 50% to classify a pair of station-program-day transcripts as exhibiting duplication. That is, we concluded there was duplication of content if 50% of their combined 5-gram shingles were found to be exact matches. This is a high (i.e., conservative) threshold for defining text reuse, implying that a pair of program transcripts must show substantial reuse within at least half of their transcribed texts for us to conclude that the pair had text reuse. However, we adopted this threshold to ensure that any duplication findings and conclusions were not affected by any lingering non-news content that may not have been fully removed from our original transcripts by TVEyes.
Although this approach is based on established methods for detecting text reuse at the pairwise document level for a corpus of documents, the total size of our 91-day station-program-day transcripts made the application of these methods computationally intractable. To see this, note that each individual pairwise text reuse comparison can take several seconds on a standard desktop computer. Implementing these pairwise comparisons for the full sample of 419,137 station-program-day transcripts would have required more than 80 billion individual pairwise comparisons, which would have required hundreds of thousands of days of computation time. To address this, we adjusted our candidate windows of text reuse comparison so that we implemented pairwise comparisons only when they fell within three days of one another using a rolling 3-day window. Over the 91 days of our sample period there were 268 such three-day windows. Although more limited, these rolling three-day time windows were a more realistic timeframe for potential news content duplication than the full 91-day time window. We examined the content of the station pairs across all 91 days, but in rolling three-day increments. This modification still yielded approximately 13,896 broadcast transcripts that resulted in more than 96 million implementations of our text reuse method for each three-day period.
What We Found
Before we discuss the fundamental questions, we present the attributes of the television markets that comprise the database for the research.
The DMA Groups
We divided the 210 DMAs into six groups: Group1 includes the top 25 markets in the country; Group2 consists of DMAs 26-50; Group3 is 51-75; Group 4 is 76-100; Group5 is 101-150 and Group 6 is 151-210. Using this framework, each of the groups comprised 25 DMAs, except for Groups 5 and 6, which accounted for 50 and 60 markets, respectively.
Two attributes of the DMA Groups provide important context: (1) size, as measured by the total number of households with a television; and (2) the total number of stations in the group.
Size: Nielsen reported that, as of January 1, 2019, there were 110,244,650 households in the U.S. with at least one television (Nielsen, 2019). We determined the size of the DMA Groups as a proportion of the total number of households with a television. DMA Group 1, the top 25 markets in the country, accounted for 50 percent of that total (See Fig 2, indicated as circles). DMA Group 2 accounted for another 18 percent. The two largest DMA groups comprised more than two-thirds of households (68%) with a television. The other four DMA Groups were significantly smaller, with DMA Group 3 at 11 percent, DMA Group 4 at 8 percent; and DMA Groups 5 and 6 at 9 and 4 percent, respectively. DMA Groups 4, 5, and 6 collectively accounted for the remaining 32 percent of households with a television.
Fig. 2: Television Households (TvHHs) and Stations in the DMA Groups
Total Number of Stations: The station database comprises 861 stations that regularly delivered news broadcasts during the Fall of 2019, the period for which we have their content. They were distributed across the 210 markets in the country and, by extension, across the DMA Groups that we defined. We report that distribution for each DMA Group as a proportion of those stations. There was less variation in that proportion across the groups than across television households.
The markets in DMA Group 1 had a combined total of 178 stations, which accounted for 21 percent of the stations in our database (N=861) (See Fig 2, as indicated by a diamond).
DMA Group 2 accounted for 123 stations, 14% of the total (Fig. 2).
DMA Groups 3 and 4 exhibited very similar profiles. Each accounted for 12 percent of the total number of stations (Fig. 2), with 106 and 104 stations, respectively.
DMA Groups 5 and 6 were considerably larger than the first four DMA Groups, with 50 and 60 markets, respectively. DMA Group 5 accounted for more than one-fifth of stations (n=189, 22%).
DMA Group 6 consisted of 60 markets with a total of 161 stations (19%, Fig 2).
CONNECTED ARRANGEMENTS:
DUOPOLIES, SERVICE AGREEMENTS & COMMON OWNERSHIP
To understand the local television landscape, we must consider the prevalence of connections among stations. Those connections take one of three forms: (1) a duopoly, in which one station group owns stations within the same DMA; (2) service agreements (SA), formal arrangements among stations to share resources (as described above); and (3) common ownership, in which one station group owns stations in different markets. We use just one term, “connected” stations, to describe all these forms.
We defined service agreements narrowly. We did not include service agreements with low-power stations with the same call letters, nor service agreements with digital or satellite stations. We examined only the service agreements that would affect the control of separate stations and that were important to consider in broadcast ownership policy.
We found that connected stations were ubiquitous; they existed in 176 (84%) out of the 210 markets in the country. Further, those 176 connected markets (CA markets) accounted for more than 87 percent of the television households in the U.S. The following analysis refers to those DMAs.
Markets within DMA Groups with Connected Station Pairs
For the DMA groups, we calculated the proportion of DMAs within each group in which at least one station was party to at least one connected arrangement (CA). In DMA Group 1, there were connected arrangements (CAs) in 22 of the 25 markets (88%) (Fig 3, as indicated by a solid circle). In DMA Group 2, 17 of 25 markets (68%) had CAs, the lowest proportion across all the groups. DMA Group 3 matched the first group with CAs in 22 of 25 markets (88%). DMA Group 4 saw 21 of 25 markets with CAs (84%). DMA Group 5 had the highest proportion of markets with connected arrangements (46 out of 50 markets (92%)). In DMA Group 6, 48 of 60 markets (80%) had connected arrangements.
Fig. 3: Connected Arrangements (CA) in the DMA Groups
Stations in DMA Groups with Connected Station Pairs
We also examined the prevalence of connected arrangements in markets by examining stations within the markets. That is, we calculated the average percentage of stations in the markets within the DMA groups that were part of a connected station pair.
We found that the first four DMA Groups registered similar percentages. In DMA Group 1, an average of 59 percent of stations in its markets were part of a connected arrangement (Fig 3, as indicated by a diamond). In DMA Groups 2, 3 and 4, the proportion was 56, 61 and 60 percent, respectively. In DMA Groups 5 and 6, about three-fourths of their stations were part of a CA (73 and 74 percent, respectively) (Fig. 3).
The trend is clear. Although connected arrangements are a common feature of the television station landscape, they are especially a feature of the smaller markets in the country.
OWNERSHIP, CONTROL & DUPLICATION
The second fundamental question referred to the nature of ownership and/or control of television stations in the U.S. and, most importantly, the possible effect of that control on broadcast content. We address these questions one at a time.
Ownership, Control & Connected Arrangements
Determining who controls a television station is not a straightforward proposition because “control” is not conferred simply by ownership. Indeed, a station owned by one entity may be “controlled” by a third through a service agreement. We showed above (See Fig. 3) that connected arrangements are common across the country. However, our concern was not just whether a connected arrangement was in place. Rather, our question was whether that arrangement affected content. To answer that question, we examined the nature of the connected arrangement. For service agreements, we defined a controlling service agreement as having a brokering station that was a content originator and a brokered station that was a non-originator. Using this approach, we defined control and content in the most conservative manner. A control relationship required two characteristics: (1) a brokering station (controller) that was an originator of content; and (2) a brokered station (controlled) that simply presented content originated by the brokering station.
It is important to note that, in two-thirds of the service agreements, the brokering and brokered stations shared the same news studio. When that is the case, control has important implications for news content.
Given these considerations, the determination of control required the application of a formula. We began with the number of stations that were owned by the station group. Then we added the number of stations that the station group controlled through a SA, in which it was the originator of content. Then we subtracted the number of stations that the station group owned, that were controlled by another entity, through a SA in which the other station group was the originator of content. Through this process, we determined the number of stations that a station group controlled among the 861 stations that comprised our sample. Table 1 shows the result of this process for the top five ownership/control groups in the sample. It is important to note that these five groups account for 52 percent (445/861) of the stations in our analysis. A complete list of owners and the number of stations they owned and controlled appears in Appendix A.
For more context, the next 13 station groups controlled another 28 percent of the stations in our sample. Added to the top five groups, just 18 station groups control 80 percent of the stations in our sample. However, the difference between the top five station groups and the rest of the owners in the sample was significant. Therefore, we use only the top five station groups here to provide a clearer picture of the conditions.
Table 1: # Stations owned/control by Top Five station groups
Even among the top five ownership/control groups, there were clearly dominant actors. Nexstar Media Group controlled the highest proportion of stations (16%, n=138) for a station group, over three times as many stations as the fifth station group, E.W. Scripps (5%, n=43), and more than twice as many as Tegna (7%, n=60). Gray Television at 14 percent (n=118) and Sinclair at 10 percent (n=86) were also significant players in news production.
Duplication
The second fundamental question we considered was whether ownership/control affected news content. We defined “affect” as the exact duplication of text (text reuse). Our text reuse methodology (see above) yielded a binary classification of whether each broadcast station pair in our sample was a news broadcast duplicator during the study period. Further, we required the station pair to be connected through one of three arrangements: a service agreement (as described above); a duopoly, in which one station group owned both stations in the same DMA (the FCC definition); or common ownership, in which the stations were in different markets but owned by the same station group. Specifically, duplication for a station pair had four conditions. Duplication therefore required that the station pair:
- Be connected through a jointly held non-digital service agreement (as defined above) or a duopoly or common ownership (owned by the same entity but the stations are in different markets);
- Have one station that is a content originator; and
- Have one member that is not a content originator; and
- At least 50% of the content of a station pair’s transcripts must be an exact match.
This definition of duplication is both very strict and conservative. In addition to the connection requirements between the stations in the duplicating station pair, we required the content to be an exact match. To put it simply, the exact match would qualify as plagiarism in another context. Further, that exact match had to occur in at least 50 percent of the broadcast. We took this approach to ensure that the duplication we found was just that, duplication.
Ownership/Control & Duplication
The level of duplication was greatly affected by the ownership/control condition of the station pairs. The analysis revealed a clear story. Across all station pairs, those that were connected as part of a service agreement exhibited significantly higher rates of duplication (51%, on average) than did station pairs that were bound by ownership (including duopoly or common ownership) (25%, on average). Figure 4 shows the results of that analysis for the top five ownership station groups (as in Table 1).
Fig. 4: Station Pairs Exhibiting Duplication
Service-Agreement-controlled station pairs exhibited much more duplication than did owned station pairs across all the station groups. For Nexstar, SA duplication (53%) was over twice as prevalent (220%) than ownership duplication (24%). Gray station pairs exhibited the smallest difference (26% to 40%), but it was substantial (65%) nevertheless. Sinclair’s SA controlled station pairs exhibited four times more duplication (36%) than its owned station pairs (9%). The difference for Tegna stations was 235 percent (14% to 33%). EW Scripps station pairs exhibited, by far, the largest difference (12% to 100%).
DMAs and Duplication
We applied the duplication findings to the DMA Groups that accounted for 210 markets. Duplication occurred in 84 (39%) of all markets. These markets accounted for almost 41 million (37%) households with a television in the U.S. at the time of our study.
There was a distinct pattern in the distribution of the 96 duplicating station pairs across the DMA Groups. The top 100 DMAs are in DMA Groups 1 to 4 (each with 25 DMAs). Collectively, they accounted for 40 percent of the duplicating stations with the highest proportion in DMA Group 1 (14%) and lowest in DMA Group 4 (7%) (Fig. 5). However, the two DMA Groups that represent the smallest television markets (DMA Groups 5 and 6) contained 110 markets collectively. They accounted for 60 percent of the duplicating station pairs (28% and 32% for DMA Groups 5 and 6, respectively). Clearly, duplicating station pairs were a substantial feature in the smallest television markets across the country.
Fig. 5: % Duplicating Station Pairs across DMA Groups
DUPLICATION IN TELEVISION MARKETS
We gathered content from the newscasts of stations in all 210 markets in the U.S. And we defined duplication quite narrowly (as outlined above): essentially the broadcast activity of station pairs in which one station controlled the other, either by ownership or through a service agreement. Further, one station was an originator of content (the broker), the other presented content (the brokered station), and the station pair exhibited at least 50 percent of duplicated content. We present the information for each television market in which we found duplication, specifically, the call letters of the station pair, the controlling entity of each station, the number of three-day windows in which we found duplication and the average proportion of content duplication exhibited by the station pair. First, some definitions and summary findings:
The Rolling 3-day Window
As we stated previously, the sheer volume of the transcripts over the 91-day sample period made the implementation of the pairwise comparisons for the full sample of 663,481 station-program-day transcripts computationally impracticable. Our solution was to adjust our candidate windows of text reuse comparison. We limited the pairwise comparisons for our program-day-level transcripts to a series of rolling three-day windows. That is, the broadcasts occurred within three days of one another. Within the 91-day sample period, there were 268 three-day time frames. For example, the first rolling 3-day period was days 1, 2, and 3; the second period was days 2, 3, and 4; the third period was days 3, 4, and 5, and so on, until we covered all 91 days. To be clear, we examined the content of the station pairs across all 91 days, but we did so in rolling three-day increments. We report the percentage of those rolling three-day time frames in which duplication occurred.
On average, the station pairs exhibited duplication at our 50 percent threshold in almost two-thirds (65%) of those three-day windows. DMA Group 2 registered the highest proportion of duplicating three-day time frames at 77 percent. That was followed by DMA Group 6 (71%); DMA Group 3 (69%); DMA Group 1 (67%); DMA Group 5 (62%). DMA Group 4 was the outlier with less than a quarter of such duplicating windows (23%).
Duplicating Station Pairs
There were 96 station pairs that exhibited duplication of content above our 50 percent threshold. There were only 182 unique stations in the station pairs, because some stations had service agreements or ownership arrangements with multiple stations.
The duplicating station pairs were distributed across all DMA Groups, with the majority in the smaller markets represented by DMA Group 6 (33%, 32 station pairs) and DMA Group 5 (28%, 27 station pairs). DMA Group 1 had just over 13 percent (13 station pairs); DMA Group 3 had just over 10 percent (10 station pairs). DMA Group 2 had just over 8 percent (8 station pairs) and DMA Group 4 had just over 7 percent (7 station pairs).
Across all station pairs, the average proportion of duplicated content was 69 percent, with a range of 51 to 96 percent. However, there was variation among the DMA Groups, and that variation was statistically significant (p=<.001). The average proportion of duplicated content was highest among the station pairs in DMA Group 1 (86%) on 65 percent of the rolling 3-day windows. That was followed by DMA Group 2 at 74 percent (on 77 percent of the rolling 3-day windows). The order after that was DMA Group 6 (69%), on 70 percent of rolling 3-day windows; DMA Group 5 (66%), on 64 percent of rolling 3-day windows; DMA Group 3 (64%), on 70 percent of the rolling 3-day windows; and DMA Group 4 (59%), on just 15 percent of the rolling 3-day windows.
Across the six groups, the average proportion of rolling 3-day windows on which duplication occurred was 64 percent. The order was: DMA Group 2 at 77 percent, followed by DMA Groups 3 and 6 at 70 percent, then DMA Groups 1 and 5 at 65 percent and, lastly, DMA Group 4 at 16 percent.
Market Profiles
There were 84 DMAs in which at least one station was part of a duplicating pair. That is, the market included either both stations of the duplicating station pair (Originating and Non-Originating station) or only the Non-Originating station of the pair. Using this approach, the findings revealed five market profiles of duplicating station pairs (Fig. 6):
Fig. 6: Market Profiles and Duplication
Market Profile 1: Markets in which there was one duplicating station pair within the DMA. There were 60 (71%) such markets.
Across these markets, the average proportion of duplication among the station pairs was 67 percent on an average of 65 percent of the rolling 3-day windows.
Market Profile 2. Markets with two duplicating station pairs within the DMA. There were eight (10%) such markets. They included:
Jacksonville, FL (#42)
Albuquerque, NM (#47)
Springfield, MO (#72)
Santa Barbara-Santa Maria-San Luis Obispo, CA (#124)
Youngstown, OH (#125)
Chico-Redding, CA (#132
Palm Springs, CA (#145)
Eureka, CA (#195)
In these markets, the average proportion of duplication among the station pairs was 67 percent across an average of 63 percent of the rolling 3-day windows.
Market Profile 3: Markets with one duplicating station pair within the DMA and one originating station that was paired with a non-originating station in another DMA. There were seven (8%) such markets:
San Francisco, CA (#8)
Boston, MA (#9)
Orlando, FL (#18)
Raleigh, NC (#25)
Boise, ID #100)
San Angelo, TX (#196)
Casper-Riverton, WY (#198)
In these markets, the average proportion of duplication among the station pairs was 77 percent across an average of 56 percent of the rolling 3-day windows.
Market Profile 4: Markets in which there was at least one originating station paired with a non-originating station in another DMA. There were five (6%) such markets:
Detroit, MI (#14)
Anchorage, AK (#147)
Missoula, MT (#164)
Great Falls, MT (#192)
Cheyenne, WY – Scottsbluff, NE (#197).
In these markets, the average proportion of duplication among the station pairs was 81 percent across an average of 89 percent of the rolling 3-day windows.
Market Profile 5: Markets that contained only the non-originating station of the duplicating pair. There were only four (5%) such markets:
Reno, NV (#109)
Abilene- Sweetwater, TX (#165)
Butte, MT (#185)
Juneau, AK (#207)
In these markets, the average proportion of duplication among the station pairs was 88 percent across an average of 99 percent of the rolling 3-day windows. To be clear, these markets reflect the “receiving” DMAs of some markets specified in Profile 4. However, we report the findings for these DMAs separately.
Duplicating Inside/Outside of the Market
An important factor in the duplicating activity of the station pairs is the extent to which the duplication occurred between stations that were inside or outside of a particular market. In a significant majority of cases, 86 percent (83 out of 96 duplicating pairs), the duplication occurred inside the market. In the other 14 percent (13/96) there was at least one station whose duplicating partner was outside the market.
There was a sizable difference in the activity of these station pairs. For the outside-of-market station pairs, the average duplication of content was 80 percent over 83 percent of the rolling 3-day windows. For markets in which the duplicating pairs were inside the DMA the average duplication of content was 68 percent on 61 percent of the rolling 3-day windows.
Fig. 7: % Duplication Inside/Outside the Markets
Control Profile of the Station Pairs
The control profile of the station pairs consisted of three categories (Fig. 8):
1. Service Agreements (SAs): 50 (52%) of the station pairs were connected through a service agreement in which the controlling entity of the originating station was also the owner of the non-originating station. On average, station pairs in this category duplicated 64 percent of content across 57 percent of the rolling 3-day windows.
2. Duopolies: 36 (38%) of the station pairs were duopolies, that is, the controlling entity owned both stations and they were in the same market. On average, station pairs in this category duplicated 73 percent of content across 69 percent of the rolling 3-day windows.
3. Common Ownership: 10 (10%) station pairs had a common owner, but the stations were in different markets. On average, station pairs in this category duplicated 81 percent of content across 84 percent of the rolling 3-day windows.
Fig. 8: Station Pair Control Profile and Duplication
Station Groups and Duplicating Station Pairs
There were 29 station groups that controlled the duplicating station pairs (Figure 9). However, that control was not evenly distributed. A small number of station groups--Nexstar, Gray, News-Press & Gazette and Sinclair combined--controlled over half (53%) of the duplicating station pairs.
Nexstar was, by far, the most active controller, responsible for controlling 21 (22%) of the pairs. These pairs, on average, duplicated 66 percent of content on 56 percent of the rolling 3-day windows. However, the averages are a bit misleading. There was wide variation among the markets in which Nexstar was the controlling entity. For example, the proportion of duplicated content ranged from 52 percent in the San Angelo, TX (#196) market to 89 percent in Norfolk, VA (#44).
Similarly, the proportion of rolling 3-day windows during which the duplication occurred ranged from 6 percent in Davenport, IA (#98) to 100 percent in both Norfolk, VA (#44) and Lansing, MI (#110). Nexstar’s most prominent partner in the station pairs was Mission Broadcasting, accounting for 10 (47%) of the non-originating stations.
The second most active controller was Gray Television. It was the controller of 16 (17%) of the duplicating pairs. On average, these pairs duplicated 69 percent of content on 75 percent of the rolling 3-day windows. The range of duplication was not as wide as Nexstar—53 percent in Grand Junction, CO (#187) to 86 percent in Clarksburg-Weston, WV (#170).
The percentage of rolling 3-day windows ranged from 6 percent in Tucson, AZ (#73) to 100 percent in six markets (Honolulu, HI (#66), Springfield, MO (#72), Wichita, KS (#76), Augusta, GA (#105), Anchorage, AK (#147), Harrisonburg, VA (#175).
The third and fourth most prominent controlling station groups were News-Press & Gazette & Company (controlling 8 pairs, 8%) and Sinclair (controlling 6 pairs, 6%). The percentages of duplication and proportion of rolling 3-day windows were 67 percent and 64 percent (News-Press & Gazette) and 69 percent and 73 percent (Sinclair).
There were no other station groups that controlled more than four of the duplicating pairs.
Fig 9: Controlling Station Groups and Number of Duplicating Station Pairs
THE DUPLICATING MARKETS
The following is a list of the markets in which there was at least one duplicating station pair or a station that was part of a duplicating station pair in which one station was in another DMA. They are grouped by the market profiles that were presented in Figure 6 and they are listed by the size of the DMA, from the largest to smallest market in that profile group. The findings show: (1) the call letters of the duplicating pair, (2) which station is the originator (Orig) and non-originator (Non-Orig) of content, (3) the controlling entity of the originating station, (4) the owner of the non-originating station, (5) the nature of that control (service agreement, duopoly or common ownership), (6) the average proportion of content that was duplicated for the station pair; and (7) the proportion of the rolling three-day time frames during which the duplication occurred.
As stated above, in the overwhelming majority of cases (86%), the duplicating station pairs were in the same market (see Fig. 7). If there was one station of the duplicating pair in another DMA, that was noted in the description of ‘market’.
It is important to note that the controlling entity in any station pair was always the controller of the originating station.
A map showing the 84 television markets in which there is at least one station of a duplicating station pair appears in Fig. 10. The map shows the DMAs by market profile as specified above (Fig. 6). The legend below matches the colors of the markets with their duplicating profile.
Fig. 10: Markets with duplicating local television station pairs in the U.S.
Tap or click any area on the map to see detailed information about that market.
DMA Duplication Profile
Map by: David Racca, Policy Scientist, Center for Applied Demography and Survey Research (CADSR), Biden School of Public Policy and Administration, University of Delaware
Market Profile 1
Markets in which there was one duplicating station pair within the DMA. There were 60 (71%) markets.
Across these markets, the average proportion of duplication among the station pairs was 67 percent on an average of 65 percent of the rolling 3-day windows.
New York, NY (#1)
In New York the duplicating pair was WCBS (Orig) and WLNY (Non-Orig). Both stations were owned by the CBS-TV station group as part of a duopoly. On average the duplication of content for the station pair was 47 percent on 30 percent of the rolling 3-day windows.
Tampa, FL (#11)
WFLA (Orig) and WTTA (Non-Orig) were both owned by Nexstar through a duopoly. They duplicated, on average, 82 percent of content on 100 percent of the rolling 3-day windows.
Denver, CO (#17)
In Denver, station KUSA (Orig) and KTVD (Non-Orig) were both owned by Tegna in a duopoly. On average, they duplicated 95 percent of content in 100 percent of the rolling 3-day windows.
Sacramento, CA (#20)
KOVR (Orig) and KMAX (Non-Orig) were the duplicating pair, each owned by CBS-TV through a duopoly. On average they duplicated 96% of content on 3 percent of the rolling 3-day windows.
Portland, OR (#22)
KPTV (Orig) and KPDX (Non-Orig) were owned by Meredith Corporation through a duopoly. On average they duplicated 91 percent of content on 84 percent of the rolling 3-day windows.
Columbus, OH (#34)
The station pair, WSYX (Orig), controlled by Sinclair, and WTTE (Non-Orig) owned by Cunningham Broadcasting, were connected through a service agreement. On average they duplicated 51 percent of content on 45 percent of the rolling 3-day windows.
Austin, TX (#40)
The duplicating pair was KXAN (Orig), controlled by Nexstar and KNVA (Non-Orig), owned by 54 Broadcasting. They were connected through a service agreement. On average they duplicated 62 percent of content on 99 percent of the rolling 3-day windows.
Norfolk, VA (#44)
WAVY (Orig) and WVBT (Non-Orig) were in a duopoly owned by Nexstar. They duplicated 63 percent of content on 99 percent of the rolling 3-day windows.
Oklahoma City, OK (#45)
KWTW (Orig) and KSBI (Non-Orig) were in a duopoly owned by Griffin Communications. On average they duplicated 93 percent of content on 90 percent of the rolling 3-day windows.
Memphis, TN (#51)
In Memphis, the originating station, WANT (Orig), was controlled by Tegna. It was connected through a service agreement with WJKT (Non-Orig), owned by Nexstar. They duplicated, on average, 64 percent content on 30 percent of the rolling 3-day windows.
Buffalo, NY (#52)
Stations WIVB (Orig) and WNLO (Non-Orig) were connected by a duopoly owned by Nexstar. They duplicated 58 percent of content on 60 percent of the rolling 3-day windows.
Providence, RI (#53)
The station pair in the market were WPRI (Orig), controlled by Nexstar and WNAC (Non-Orig), owned by Mission Broadcasting. They were connected by a service agreement, and they duplicated 58 percent of content on 94 percent of the rolling 3-day windows.
Albany, NY (#59)
The station pair in Albany was WTEN (Orig), controlled by Nexstar and WXXA (Non-Orig), owned by Mission Broadcasting. They were connected by a service agreement. They duplicated 53% of content on 33 percent of the rolling 3-day windows.
Wilkes-Barre, PA (#62)
The station pair in Wilkes-Barre was WBRE (Orig), controlled by Nexstar and WYOU (Non-Orig), owned by Mission Broadcasting. They were connected through a service agreement. They duplicated 71 percent of content on 100 percent of the rolling 3-day windows.
Dayton, OH (#64)
WDTN (Orig) was controlled by Nexstar, and it was connected to WBDT (Non-Orig) owned by Vaughn Media LLC by a service agreement. They duplicated 56 percent of content on 78 percent of the rolling 3-day windows.
Honolulu, HI (#66)
KGMB (Orig) and KHNL (Non-Orig) were both controlled by Gray in a duopoly. They duplicated 63 percent of content on 100 percent of the rolling 3-day windows.
Charleston, WV (#70)
The station pair was WCHS (Orig), controlled by Sinclair, and WVAH (Non-Orig), owned by Cunningham Broadcasting. They were connected through a service agreement. They duplicated 84 percent of content on 100 percent of the rolling 3-day windows.
Tucson, AZ (#73)
KOLD (Orig), controlled by Gray, and KMSB (Non-Orig), owned by Tegna, were connected through a service agreement. They duplicated 54 percent of content on 24 percent of the rolling 3-day windows.
Wichita, KS (#76)
KWCH (Orig) and KSCW (Non-Orig) were both controlled by Gray through a duopoly. They duplicated 52 percent of content on 15 percent of the rolling 3-day windows.
Syracuse, NY (#81)
WSTM (Orig), controlled by Sinclair and WTVH (Non-Orig), owned by Granite Broadcasting were connected through a service agreement. They duplicated 78 percent of content on 38 percent of the rolling 3-day windows.
Shreveport, LA (#90)
KTAL (Orig), controlled by Nexstar and KMSS (Non-Orig), owned by Mission Broadcasting were connected by a service agreement. They duplicated 55 percent of content on 14 percent of the rolling 3-day windows.
Jackson, MS (#92)
WLBT (Orig), controlled by Gray and WDBD (Non-Orig), owned by American Spirit Media were connected through a service agreement. They duplicated 58 percent of content on 6 percent of the rolling 3-day windows.
Burlington, VT (#96)
WFFF (Orig), controlled by Nexstar and WVNY (Non-Orig), owned by Mission Broadcasting were connected through a service agreement. They duplicated 60 percent of content on 37 percent of rolling 3-day windows.
Davenport, IA (#98)
WHBF (Orig) was controlled by Nexstar and it was in a service agreement with KLJB (Non-Orig), owned by Mission Broadcasting. They duplicated 51 percent of content on 5 percent of the rolling 3-day windows.
Ft. Smith-Fayetteville, AR (#101)
KNWA (Orig) and KFTA (Non-Orig) were both controlled by Nexstar through a duopoly. They duplicated 55 percent of content on 24 percent of the rolling 3-day windows.
Augusta, GA (#105)
WRDW (Orig) and WAGT (Non-Orig) were both controlled and owned by Gray through a duopoly. They duplicated 87 percent of content on 96 percent of the rolling 3-day windows.
Springfield-Holyoke, MA (#108)
WGGB (Orig) and WSHM (Non-Orig) were controlled by Meredith Corp through a duopoly. They duplicated 82 percent of content on 100 percent of the rolling 3-day windows.
Lansing, MI (#110)
WLNS (Orig), controlled by Nexstar and WLAJ (Non-Orig), owned by Mission Broadcasting, were connected through a service agreement. They duplicated 72 percent of content on 96 percent of the rolling 3-day windows.
Montgomery, AL (#116)
WAKA (Orig), controlled by Bahakel Communications Limited, and WNCF (Non-Orig), owned by SagamoreHill Broadcasting were connected through a service agreement. They duplicated 55 percent of content on 93 percent of the rolling 3-day windows.
Fargo, ND (#117)
KVLY (Orig) and KXJB (Non-Orig) were both controlled by Gray in a duopoly. They duplicated 62 percent of content on 99 percent of the rolling 3-day windows.
Traverse City, MI (#120)
WWTV (Orig) was controlled by Heritage Broadcasting Group and was connected through a service agreement to WFQX (Non-Orig), owned by Cadillac Telecasting Co. They duplicated 53 percent of content on 98 percent of the rolling 3-day windows.
Lafayette, LA (#121)
KADN (Orig) and KLAF (Non-Orig) were both owned by Allen Media Broadcasting LLC through a duopoly. They duplicated 78 percent of content on 100 percent of the rolling 3-day windows.
Eugene, OR (#123)
KVAL (Orig), controlled by Sinclair and KLSR (Non-Orig), owned by California-Oregon Broadcasting were connected through a service agreement. They duplicated 91 percent of content on 99 percent of the rolling 3-day windows.
Corpus Christi, TX (#128)
KRIS (Orig), controlled by E.W. Scripps and KZTV (Non-Orig), owned by SagamoreHill Broadcasting, were connected through a service agreement. They duplicated 57 percent of content on 55 percent of the rolling 3-day windows.
Wilmington, NC (#129)
WECT (Orig), controlled by Gray and WSFX (Non-Orig), owned by American Spirit Media LLC were connected through a service agreement. They duplicated 56 percent of content on 36 percent of the rolling 3-day windows.
Amarillo, TX (#131)
KAMR (Orig) was controlled by Nexstar and it was connected to KCIT (Non-Orig), owned by Mission Broadcasting, through a service agreement. They duplicated 54 percent of content on 22 percent of the rolling 3-day windows.
Columbus-Tupelo-West Point, MS (#133)
WTVA (Orig), controlled by Allen Media Broadcasting LLC, and WLOV (Non-Orig), owned by Coastal Television Broadcasting LLC were connected through a service agreement. They duplicated 66 percent of content on 96 percent of the rolling 3-day windows.
Medford-Klamath, OR (#135)
KOBI, controlled by California Oregon Broadcasting, and KMVU, owned by Terrier Media Holdings, were connected through a service agreement. They duplicated 51 percent of content on 3 percent of the rolling 3-day windows.
Columbia-Jefferson City, MO (#136)
KMIZ (Orig) and KQFX (Non-Orig) were both owned by News-Press & Gazette and connected through a duopoly. They duplicated 51 percent of content on 15 percent of the rolling 3-day windows.
Monroe, LA – El Dorado, AR (#137)
KTVE (Orig) controlled by Mission Broadcasting and KARD (Non-Orig) owned by Nexstar were connected by a service agreement. They duplicated 69 percent of content on 29 percent of the rolling 3-day windows.
Rockford, IL (#139)
WTVO (Orig), controlled by Nexstar, and WQRF (Non-Orig), owned by Mission Broadcasting were connected through a service agreement. The station pair duplicated 59% of content on 76 percent of the rolling 3-day windows.
Topeka, KS (#141)
KSNT (Orig), controlled by Nexstar and KTKA (Non-Orig), owned by Vaughn Media LLC, were connected through a service agreement. They duplicated 66 percent of content on 100 percent of the rolling 3-day windows.
Erie, PA (#151)
WICU (Orig), controlled by SJL Broadcast Management Group, and WSEE (Non-Orig), owned by Lilly Broadcasting were connected through a service agreement. They duplicated 78 percent of content on 100 percent of the rolling 3-day windows.
Joplin, MO – Pittsburg, KS (#153)
KOAM (Orig), controlled by Morgan Murphy Media and KFIX (Non-Orig), owned by SagamoreHill Midwest LLC, were connected through a service agreement. The station pair duplicated 74 percent of content on 3 percent of the rolling 3-day windows.
Bangor, ME (#155)
WVII (Orig), controlled by Bangor Communications and WFVX (Non-Orig), owned by SDR Rockfleet Holdings LLC, were connected through a service agreement. They duplicated 58 percent of content on 78 percent of the rolling 3-day windows.
Gainesville, FL (#157)
WGFL (Orig), controlled by CPC Media LLC and WNBW (Non-Orig), owned by MPS Media, were connected through a service agreement. They duplicated 68 percent of content on 93 percent of the rolling 3-day windows.
In addition, as stated in the information for the Orlando DMA, Gainesville station WOGX (Non-Orig) was connected to Orlando station WOFL (Orig). Both were owned by Fox. On average they duplicated 92 percent of content on 92 percent of the rolling 3-day windows.
Binghamton, NY (#160)
WIVT (Orig) and WBGH (Non-Orig) were both owned by Nexstar as a duopoly. They duplicated 80 percent of content on 100 percent of the rolling 3-day windows.
Idaho Falls – Pocatello, ID (#161)
KIFI (Orig), controlled by News-Press & Gazette Co and KIDK (Non-orig), owned by VistaWest Media LLC, were connected through a service agreement. They duplicated 66 percent of content on 97 percent of the rolling 3-day windows.
Yuma, AZ – El Centro, CA (#166)
KECY (Orig), controlled by News-Press & Gazette Company and KYMA (Non-Orig), owned by Terrier Media Holdings were connected through a service agreement. They duplicated 60 percent of content on 73 percent of the rolling 3-day windows.
Utica, NY (#169)
WUTR (Orig), controlled by Mission Broadcasting and WFXV (Non-Orig), owned by Nexstar were connected through a service agreement. They duplicated 52 percent of content on 19 percent of the rolling 3-day windows.
Clarksburg – Weston, WV (#170)
WDTV (Orig) and WVFX (Non-Orig) were connected by a duopoly controlled by Gray. They duplicated 86 percent of content on 75 percent of the rolling 3-day windows.
Dothan, AL (#173)
WTVY (Orig) and WRGX (Non-Orig) were both owned by Gray in a duopoly. They duplicated 65 percent of content on 78 percent of the rolling 3-day windows.
Harrisonburg, VA (#175)
WHSV (Orig) and WSVF (Non-Orig) were both owned by Gray in a duopoly. The station pair duplicated 71 percent of content on 100 percent of the rolling 3-day windows.
Watertown, NY (#178)
WWNY (Orig) and WNYF (Non-Orig) were controlled by Gray though a duopoly. They duplicated 55 percent of content on 56 percent of the rolling 3-day windows.
Charlottesville, VA (#183)
WCAV (Orig), controlled by Lockwood Broadcasting was connected by a service agreement with WVAW (Non-Orig), owned by Standard Media Group LLC. They duplicated 83 percent of content on 80 percent of the rolling 3-day windows.
Bend, OR (#186)
KTVZ (Orig) and KFXO (Non-Orig) were owned by News-Press & Gazette Company in a duopoly. They duplicated 59 percent of content on 37 percent of the rolling 3-day windows.
Grand Junction – Montrose, CO (#187)
KKCO (Orig) and KJCT (Non-Orig) were both owned by Gray in a duopoly. The duplicated 53 percent of content on 81 percent of the rolling 3-day windows.
Twin Falls, ID (#189)
KMVT (Orig) and KSVT (Non-Orig) were both by Gray in a duopoly. They duplicated 65 percent of content on 83 percent of the rolling 3-day windows.
Meridian, MS (#191)
WGBC (Orig), controlled by Standard Media Group LLC and WMDN (Non-Orig) owned by Sheldon Galloway were connected through a service agreement. They duplicated 66 percent of content on 83 percent of the rolling 3-day windows.
Greenwood – Greenville, MS (#193)
WXVT (Orig) and WABG (Non-Orig) were both owned by Terrier Media Holdings in a duopoly. They duplicated 66 percent of content on 100 percent of the rolling 3-day windows.
Market Profile 2
Markets with two duplicating station pairs within the DMA, eight (10%) markets.
In these markets, the average proportion of duplication among the station pairs was 67 percent across an average of 63 percent of the rolling 3-day windows.
Jacksonville, FL (#42)
There were two duplicating pairs in the market. WJAX (Orig), controlled by Hoffman Communications and WFOX (Non-Orig), owned by Cox Media Group, were connected through a service agreement. They duplicated 90 percent of content on 100 percent of the rolling 3-day windows.
WTLV (Orig) and WJXX (Non-Orig) were in a duopoly owned by Tegna. The pair duplicated 95 percent of content on 100 percent of the rolling 3-day windows.
Albuquerque, NM (#47)
There were two duplicating pairs in the market, KLUZ (Orig), controlled by Univision and KTFQ (Non-Orig), owned by Entravision were connected through a service agreement. On average they duplicated 59 percent of content on 71 percent of the rolling 3-day windows.
The station pair of KRQE (Orig), controlled by Nexstar and KWBQ (Non-Orig), owned by Mission Broadcasting were connected through a service agreement. They duplicated 80 percent of content on 53 percent of the rolling 3-day windows.
Springfield, MO (#72)
There were two duplicating station pairs in the market. KYTV (Orig) and KSPR (Non-Orig) were both controlled by Gray in a duopoly. They duplicated 80 percent of content on 100 percent of the rolling 3-day windows.
KOLR (Orig), controlled by Mission Broadcasting and KOZL (Non-Orig), owned by Nexstar were connected through a service agreement. They duplicated 61 percent of content on 60 percent of the rolling 3-day windows.
Santa Barbara-Santa Maria-San Luis Obispo, CA (#124)
There were two duplicating station pairs in the market. KEYT (Orig) and KKFX were both owned by News-Press & Gazette Company and they were connected through a duopoly. They duplicated 64 percent of content on 8 percent of the rolling 3-day windows.
KEYT was also in a service agreement with KCOY (Non-Orig), owned by VistaWest Media LLC. They duplicated 81 percent of content on 100 percent of the rolling 3-day windows.
Youngstown, OH (#125)
There were two duplicating station pairs in the market. WKBN (Orig) and WYFX (Non-Orig) were both owned by Nexstar in a duopoly. They duplicated 54 Percent of content on 23 percent of the rolling 3-day windows.
WKBN was also connected to WYTV (Non-Orig), owned by Vaughn Media LLC, through a service agreement. They duplicated 53 percent of content on 1 percent of the rolling 3-day windows.
Chico-Redding, CA (#132)
There were two duplicating station pairs in the market connected by service agreements. One pair was KHSL (Orig), controlled by Allen Management Broadcasting LLC and KNVN (Non-Orig), owned by Maxair Media LLC. They duplicated 88 percent of content on 100 percent of the rolling 3-day windows.
KRCR (Orig), controlled by Sinclair and KCVU (Non-Orig), owned by Cunningham Broadcasting were the second service agreement. They duplicated 63 percent of content on 80 percent of the rolling 3-day windows.
Palm Springs, CA (#145)
KESQ (Orig) originated content through duopolies (technically the only triopoly among the stations) with two stations (KDFX (Non-Orig) and KPSP (Non-Orig). All the stations were owned by News Press & Gazette Co.
The KESQ/KDFX station pair duplicated 64 percent of content on 97 percent of the rolling 3-day windows.
The KESQ/KPSP station pair duplicated 88 percent of content on 100 percent of the rolling 3-day windows.
Eureka, CA (#195)
There were two duplicating station pairs in the market. KAEF (Orig) was controlled by Sinclair and KBVU (Non-Orig) was owned by Cunningham Broadcasting. They were connected through a service agreement, and they duplicated 63 percent of content on 59 percent of the rolling 3-day windows.
KIEM (Orig) and KVIQ (Non-Orig) were both owned by Terrier Media Holdings through a duopoly. They duplicated 54 percent of content on 2 percent of the rolling 3-day windows.
Market Profile 3
Markets with one duplicating station pair within the DMA and one originating station that was paired with a non-originating station in another DMA, seven (8%) markets.
In these markets, the average proportion of duplication among the station pairs was 77 percent across an average of 56 percent of the rolling 3-day windows.
San Francisco, CA (#8)
In the San Francisco market there were two duplicating station pairs. KPIX (Orig) and KBCW (Non-Orig), both owned by CBS-TV station group through a duopoly. The average proportion of duplicated content was 94 percent on 15 percent of the rolling 3-day windows.
The second duplicating pair was KTVU (Orig), controlled by Fox and KRXI (Non-Orig), owned by Cunningham Broadcasting. They were connected through a service agreement. However, while KTVU is in the San Francisco market, KRXI is in the Reno DMA (#109). The average duplicated content between the stations was 87 percent on 100 percent of the rolling 3-day windows.
Boston, MA (#9)
In Boston there were two duplicating station pairs. WHDH (Orig) and WLVI (Non-Orig), both owned by Sunbeam Television Corporation through a duopoly. They duplicated on average 92 percent of content on 100 percent of the rolling 3-day windows.
The second pair was WBZ (Orig) controlled by CBS-TV and WSHM (Non-Orig), owned by Meredith Corporation. They were connected through a service agreement. WBZ is in the Boston market while WSHM is in the Springfield-Holyoke DMA (#108). On average the station pair duplicated 94 percent of content on 15 percent of the rolling 3-day windows.
Orlando, FL (#18)
In Orlando, there were two duplicating station pairs. WESH (Orig) and WKCF (Non-Orig) were both owned by Hearst through a duopoly. On average, they duplicated 61 percent of content on 87 percent of the rolling 3-day windows.
The other station pair involved WOFL (Orig) and WOGX (Non-Orig). Both were owned by Fox, but the arrangement was not a duopoly because WOGX is in the Gainesville, FL DMA (#157). They were connected through common ownership. On average they duplicated 92 percent of content on 92 percent of the rolling 3-day windows.
Raleigh-Durham, NC (#25)
WRAL was the originating partner for two non-originating stations, WRAZ in the same market and WILM in Wilmington, NC. All the stations were controlled by Capitol Broadcasting, through a duopoly (WRAL/WRAZ) and common ownership (WRAL/WILM).
The WRAL/WRAZ station pair duplicated on average 88 percent of content on 98 percent of the rolling 3-day windows. The WRAL/WILM combination duplicated 77 percent of content on 99 percent of the rolling 3-day windows.
Boise, ID (#100)
KIVI (Orig), controlled by E.W. Scripps, and KNIN (Non-Orig), owned by Gray, were connected through a service agreement. They duplicated 57 percent of content on 6 percent of the rolling 3-day windows.
However, KIVI was in a common ownership arrangement with KSAW (Non-Orig), also owned by E. W. Scripps. KSAW is in the Twin Falls, ID market (#189). They duplicated 76 percent of content on 100 percent of the rolling 3-day windows.
San Angelo, TX (#196)
KLST (Orig), controlled by Nexstar and KSAN (Non-Orig), owned by Mission Broadcasting were connected through a service agreement. They duplicated 52 percent of content on 1 percent of the rolling 3-day windows.
As we stated in the information for the Abilene-Stillwater, TX DMA (#165) there was another duplicating station pair in which a San Angelo station was involved. Both stations were controlled by Tegna, in which the originating station, KIDY (Orig) was in the San Angelo market, but non-originating station KXVA (Non-Orig) was in the Abilene-Sweetwater market. That station pair duplicated 94 percent of content on 100 percent of the rolling 3-day windows.
Casper-Riverton, WY (#198)
KTWO (Orig), controlled by Vision Alaska, was connected to KFNB (Non-Orig), owned by Coastal Television Broadcasting LLC, through a service agreement. They duplicated 65 percent of the content on 51 percent of the rolling 3-day windows.
To reiterate what we stated in the description of the Cheyenne market, KTWO (Orig) was also the originating station for KLWY (Non-Orig) in the Cheyenne DMA.
Market Profile 4
Markets in which there was at least one originating station paired with a non-originating station in another DMA, five (6%) markets.
In these markets, the average proportion of duplication among the station pairs was 81 percent across an average of 89 percent of the rolling 3-day windows.
Detroit, MI (#14)
Detroit station WXYZ (Orig), controlled by E.W. Scripps, and WSYM (Non-Orig), owned by Gray were connected through a service agreement. WSYM, however, is in the Lansing, MI market (#110). On average they duplicated 52 percent of content on 51 percent of the rolling 3-day windows.
Anchorage, AK (#147)
The stations in Anchorage that were part of duplicating station pairs had their non-originating partners in the Juneau, AK (#207) market. Therefore, they were not duopolies. Rather they were bound by common ownership.
KYUR (Orig) and KJUD (Non-Orig) were both owned by Vision Alaska. They duplicated 96 percent of content on 100 percent of the rolling 3-day windows.
KTUU (Orig), in Anchorage, and KATH (Non-Orig), in Juneau were both owned by Gray. They duplicated 75 percent of content on 100 percent of the rolling 3-day windows.
Missoula, MT (#164)
The duplicating pair in Missoula has the originating station, KTMF (Orig) in the market and the non-originating station, KWYB (Non-Orig) in the Butte, MT (#185) DMA. Both were owned and controlled by Cowles Publishing Company. They are part of a broadcast consortium called Montana Right Now that produces local news in markets across the state. The KTMF/KWYB pair duplicated 88 percent of content on 100 percent of the rolling 3-day windows.
Great Falls, MT (#192)
The duplicating pair in Great Falls has the originating station, KFBB (Orig) in the market and the non-originating station, KWYB (Non-Orig) in the Butte, MT (#185) DMA. Both are owned and controlled by Cowles Publishing Company. They are part of a broadcast consortium called Montana Right Now that produces local news in markets across the state. The KFBB/KWYB pair duplicated 88 percent of content on 99 percent of the rolling 3-day windows.
Cheyenne, WY – Scottsbluff, NE (#197)
The stations in the Cheyenne duplicating pair stretch across two markets. They are both controlled by Gray through common ownership. KGWN (Orig) is in Cheyenne; KCWY (Non-Orig) is in the Casper-Riverton, WY (#198) market. The station pair duplicated 63 percent of content on 40 percent of the rolling 3-day windows.
There is a station pair in which the non-originating station, KLWY (Non-Orig), is in the Cheyenne market. It was owned by Coastal Television Broadcasting LLC. Its originating station was KTWO (Orig), controlled by Vision Alaska, which is in the Casper-Riverton, WY DMA. #198). The station pair was connected through a service agreement. They duplicated 65 percent of content on 56 percent of the rolling 3-day windows.
Market Profile 5
Markets that contained only the non-originating station of the duplicating pair, four (5%) markets.
In these markets, the average proportion of duplication among the station pairs was 88 percent across an average of 99 percent of the rolling 3-day windows. To be clear, these markets reflect the “receiving” DMAs of the some of the markets specified in Profile 4. However, we wanted to report the findings for these DMAs separately.
Reno, NV (#109)
As we stated in the description of the San Francisco (#8) DMA, the originating station in that market, KTVU (Orig), controlled by Fox, was connected through a service agreement to KRXI (Non-Orig), owned by Cunningham Broadcasting, in the Reno market. The average duplicated content between the stations was 87 percent on 100 percent of the rolling 3-day windows.
Abilene – Stillwater, TX (#165)
KXVA (Non-Orig) was a non-originating station in the Abilene market whose originating partner was KIDY (Orig) in the San Angelo, TX (#196) market. Both were owned by Tegna. The station pair duplicated 94 percent of content on 100 percent of the rolling 3-day windows.
Butte, MT (#185)
KWYB (Non-Orig) is a non-originating station controlled by Cowles Publishing Company. It received content from another Cowles station, KFBB (Orig) in Great Falls, MT through common ownership. The station pair duplicated 80 percent of content on 99 percent of the rolling 3-day windows.
KWYB is part of the Montana Right Now news consortium initiated by Cowles Publishing which owns and controls both stations. It also received content from KTMF (Orig) in Missoula, MT (#164) through that consortium. That pair duplicated 88 percent of content on 100 percent of the rolling 3-day windows.
Juneau, AK (#207)
Unlike the other markets, Juneau was the only market in which the stations were the non-originating partners of stations in another market—Anchorage, AK (#147), as we stated above.
KYUR (Orig) in Anchorage and KJUD (Non-Orig) in Juneau were controlled by Vision Alaska through common ownership. They duplicated 96 percent of content on 100 percent of the rolling 3-day windows.
KTUU (Orig), in Anchorage, and KATH (Non-Orig), in Juneau were both controlled by Gray through common ownership. They duplicated 75 percent of content on 100 percent of the rolling 3-day windows.
The duplicating activity of the station pairs was clearly and consistently affected by the connection between them, whether it was a service agreement, duopoly or common ownership. This listing of the markets in which that duplication occurred specifically confirms which stations controlled content for those stations that duplicated it.
CONCLUSION
There are four points in our conclusion. One refers to the methodology we used; the second addresses the business of news; the third addresses the on-going policy debate regarding broadcast ownership and the fourth examines why it matters.
The Method
It is crucial to understand that the methodology, particularly the text-reuse approach that we employed, has never been applied to the content analysis of local television news at this scale. It is a cutting-edge approach to understanding the behavior and performance of local television news stations that allows us to examine a huge corpus of local TV news content that, until now, has been unattainable. That is vitally important because the duplication of local television news content has been the subject of intense political debate. The media industry argues, first, that it does not occur in large measure. Secondly, the industry argues that, even if it does occur, it is an appropriate way to fend off the intrusion of the Internet into the news space. Our research is the most comprehensive and detailed examination of the issue.
The Business of News
The inescapable conclusion that we draw is that ownership—or more accurately control—matters in the production of local television news. The control of television stations that is derived from duopolies, service agreements, and common ownership often results in duplicated content that serves the calculus of the economies of scale. We should not be surprised by that finding. The media system is designed that way. We use private means (media firms) to deliver a public good (the news). Further, the continuing consolidation of the local television news space and the economies of scale it incentivizes make the duplication effect almost a given. Station groups are in the business to make money. They make that very point every time they advance a plan to control more stations. We do not take them to task for their profit-seeking behavior. They are private firms with a fiduciary responsibility to their shareholders. The problem is that the “commodity” through which they realize those profits is the news.
Text reuse—the duplication of the exact same material across news broadcasts--is a direct and unambiguous form of the achievement of economies of scale. The station group bears the cost of production of the story once and sells it to advertisers multiple times. It has all the hallmarks of a public good—its consumption by one consumer does not diminish its supply for another. As fewer station groups control more of the local television ecosystem, accomplishing that duplication becomes easier. And all the incentives for it are clear. Our analysis shows that those incentives are significantly utilized.
At present, the mechanisms to implement text reuse rely on the repeating of the content by live news anchors and reporters. However, there is a technology that will change the game: artificial intelligence. In July 2023, OpenAI, the Microsoft company behind ChatGPT, announced an agreement to give the American Journalism Project (AJP) $5 million “to assess the applications of AI within the local news sector” (Tameez, 2023, p. 1). AJP will tackle the thorny issues that AI presents for journalism. However, the application of AI to news dissemination has taken the form of AI-generated news anchors. In 2019, the Chinese debuted two AI-generated male news anchors modeled after two human reporters. The AI-generated reporters sound and look like their human counterparts (Zhao, 2020). The advantages of these “anchors” for the economies of scale are enormous: no days off, no rest required, no salary disagreements. As the possibilities and the incentives for duplication increase in the local television news space, the potential use of artificial intelligence to deliver that duplicated content may become irresistible.
The Policy Debate
The Federal Communications Commission (FCC) is engaged in its Quadrennial Review of Broadcast Ownership. It issued a decision in December 2023 regarding the overdue 2018 Review that maintained ownership limits and regulations. The vote was 3-2 along party lines, with the Democratic members voting for the limits and the Republican members voting against them.
There are significant disagreements with the FCC’s decision. The National Association of Broadcasters was not pleased with the decision, and the two Republican Commissioners each wrote separate dissents.
The debate will continue, as the FCC must now consider the delayed 2022 Quadrennial Review. The arguments will be the same. The broadcast industry maintains that the many sources of news and information in the new communication reality make the ownership limits obsolete. Media reformers and the FCC majority during the Biden administration, argue that, despite the communications system, none of those sources can substitute for broadcast television’s service of the agency’s regulating principles of competition, diversity, and localism. This research provides crucial information for the debate.
The position of the FCC has changed drastically since the 2024 presidential election. The new FCC Chair has forcefully stated his intention to deregulate media ownership. In the first week after the election, broadcast industry executives expressed their hopes about a new deregulatory FCC (Stilson, 2024) (Keys, 2024). The CEO of Nexstar, the largest station group and the most prominent consolidator in this research, stated, “Obviously, the number one legislative priority of Nexstar and our trade association, the NAB, is the deregulation of ownership at both the national and the local level” (Weprin, 2024, p. 1). Nexstar has gone further in its attempt to influence the FCC’s decisions. In early April 2025 it ordered most of its 160+ owned stations to air segments during local newscasts that encourage viewers to contact the FCC to demand broadcast deregulation. Most stories do not mention broadcast ownership rules or Nexstar’s position on the issue. However, they provide a link to a Nexstar-owned website. Viewers can choose pre-written social media posts which can be transmitted to the FCC with the push of a button (Keys, 2025).
Some members of Congress directly entered the policy debate in March 2025 when a bi-partisan group of 73 of them sent a letter to the FCC urging it to eliminate “antiquated ownership restrictions” (Miller, 2025).
Why Does it Matter?
The obvious question is, why does all this local news business matter? In a nutshell, it matters because our world is mediated. The narrative that the media constructs about society is crucial to how we engage with it.
The vast majority of what we know about the world every day occurs beyond our direct experience. Our knowledge of the places in which we live is constructed by people, groups, and institutions whose portrayals we accept as accurate to one extent or another. We cannot escape the condition. However, the information-seeking behavior of the public is circumscribed by the media environment that surrounds it. Citizens cannot, of their own devices, discern the rules of the game, the substance of politics or people and parties, without a mediating system to gather and disseminate information. The news matters because it is the ”hard-wiring” of our democracy (Hargreaves & Thomas, 2002, p. 4). That is particularly true in local places, and when it is diminished or absent altogether, its effect is immediate and consequential (Napoli, Weber, McCollough, & Wang, 2018) (Peterson, 2017) (Snyder & Stromberg, 2010) (Arceneaux, Dunaway, Johnson, & Vander Wielen, 2019).
Citizens use the news and, by and large, they trust it, although that trust has seriously declined over the last two decades. But that trust affects how we function as citizens. There is a direct and unambiguous connection between what we know as citizens and our engagement in the political process (Filla & Johnson, 2010) (Hayes & Lawless, 2018) (Barthel et al., 2016). There is some evidence that local television news might revamp its approach to the informational needs of communities in the face of the digital revolution and the decline of newspapers, but there are significant hurdles (Patterson, 2025).
Duplication is based on economies of scale not on the information needs of citizens.
As media firms use their stations to simply duplicate content, whether in the same market or in another market, the opportunities for local journalism decrease. That is an inescapable certainty because the news selection process is a zero-sum game. If one story is in, then another is out. Duplication is based on economies of scale, not on the public information needs of citizens. The media industry maintains that duplication is not a prominent feature of their performance. This research clearly demonstrates that duplication is not only significant but that it is accomplished through mechanisms of duopolies, common ownership, and service agreements that are specifically designed to achieve those economies of scale.
About the authors
Danilo Yanich is a professor in the Biden School of Public Policy & Administration at the University of Delaware. He has focused on the intersection of media, citizenship and public policy for two decades. His book, Buying Reality: Political Ads, Money & Local TV News (Fordham University Press, 2020), chronicles the stark differences in political ads, money, and media coverage of the presidential and down ballot campaigns in the crucial 2016 election.
Benjamin E. Bagozzi is a professor in the Department of Political Science & International Relations at the University of Delaware with a primary specialization in political methodology. He also leads the Social Analytics Data Lab and is an associate director of the Data Science Institute at UD.
Arceneaux, K., Dunaway, J., Johnson, M., & Vander Wielen, R. J. (2019). Strategic Candidate Entry and Congressional Elections in the Era of Fox News. American Journal of Political Science, 0(0). doi:10.1111/ajps.12478
Barthel, M., Holcomb, J., Mahone, J., & Mitchell, A. (2016). Civic engagement strongly tied to local news habits. Retrieved from Washington, D.C.: http://www.journalism.org/2016/11/03/civic-engagement-strongly-tied-to-local- news-habits/
Eck, K. (2025). LA Stations Saw Audience Increase During Wildfire Coverage. Retrieved from https://www.adweek.com/tvspy/la-stations-saw-audience-increase-during- wildfire-coverage/
Filla, J., & Johnson, M. (2010). Local News Outlets and Political Participation. Urban Affairs Review, 45(5), 679-692.
Flores, E., Barrón-Cedeño, A., Moreno, L., & Rosso, P. (2015). Uncovering source code reuse in large-scale academic environments. Computer Applications in Engineering Education, 23(3), 383-390.
Geiger, A. (2019). Key findings about the online news landscape in America. Retrieved from https://www.pewresearch.org/fact-tank/2019/09/11/key-findings-about-the- online-news-landscape-in-america/
Guess, A., Nyham, B., & Reifler, J. (2018). All media trust is local? Findings from the 2018 Poynter Media Trust Survey. Retrieved from St. Petersburg, FL: http://www- personal.umich.edu/~bnyhan/media-trust-report-2018.pdf
Hargreaves, I., & Thomas, J. (2002). New news, old news: Broadcasting Standards Commission.
Hayes, D., & Lawless, J. (2018). The Decline of Local News and Its Egects: New Evidence from Longitudinal Data. The Journal of Politics, 80(1), 332-336. doi:10.1086/694105
Hutchinson, A. (2020). New Report Shows Universal Distrust in Social Media as a News Source. Retrieved from https://www.socialmediatoday.com/news/new-report- shows-universal-distrust-in-social-media-as-a-news-source/571512/
Johnson, T. (2021). February Ratings/ MSNBC Tops In Total Day Viewers. Retrieved from https://deadline.com/2021/03/february-ratings-msnbc-tops-in-total-day-viewers- cnn-wins-key-demo-fox-news-takes-primetime-1234704960/
Keys, M. (2024). Broadcasters, telecoms get major win with apparent Trump victory. Retrieved from https://thedesk.net/2024/11/trump-victory-major-win- broadcasters/
Keys, M. (2025). Nexstar pushes news viewers to encourage broadcast deregulation. Retrieved from https://thedesk.net/2025/04/nexstar-fcc-must-runs-ownership- caps-deregulation/
Knight Foundation, & Gallup. (2022). American Views 2022 pt1: News in America: Public Good or Private Enterprise? Retrieved from https://knightfoundation.org/wp- content/uploads/2022/10/American-Views-2022-pt1.pdf
Leichtman Research Group. (2023). Pay TV penetration rate in the United States from 2005 to 2023. Retrieved from https://www.statista.com/statistics/467842/pay-tv- penetration-rate-usa/#statisticContainer
Lyon, C., Barrett, R., & Malcolm, J. (2006). Plagiarism is easy, but also easy to detect. Plagiary.
Lyon, C., Malcolm, J., & Dickerson, B. (2001). Detecting short passages of similar text in large document collections. Paper presented at the Proceedings of the 2001 conference on empirical methods in natural language processing.
Metzger, Z. (2024). The State of Local News: The 2024 Report. Retrieved from https://localnewsinitiative.northwestern.edu/projects/state-of-local- news/2024/report/
Miller, M. K. (2025). 73 Members Of Congress Call On FCC To Modernize Ownership Rules.
Retrieved from https://tvnewscheck.com/regulation/article/73-members-of- congress-call-on-fcc-to-modernize-ownership-rules/
Morin, A., Vasek, M., & Moore, T. (2021). Detecting Text Reuse in Cryptocurrency Whitepapers. Paper presented at the 2021 IEEE International Conference on Blockchain and Cryptocurrency (ICBC).
Napoli, P., Weber, M., McCollough, K., & Wang, Q. (2018). Assessing local journalism: news deserts, journalism divides and the determinants of the robustness of local news. Retrieved from Durham, NC:
Nielsen. (2019). Local Television Market Universe Estimates. Retrieved from https://thevab.com/storage/app/media/Toolkit/DMA_Ranker_2019.pdf
Nielsen. (2020a). In the “new normal” of COVID-19, local TV news proves to be the medium if choice for news and information. Retrieved from https://www.nielsen.com/us/en/insights/article/2020/in-the-new-normal-of-covid- 19-local-tv-news-proves-to-be-the-medium-of-choice-for-news-and-information/
Nielsen. (2020b). Local Watch Report. Retrieved from https://www.nielsen.com/us/en/insights/report/2020/nielsen-local-watch-report- diverse-young-informed-local-news-audience/
Papper, B. (2021). 2021 RTDNA/Newhouse School at Syracuse University Newsroom Survey: Local News Business Trends and Threats. Retrieved from https://assets-
002.noviams.com/novi-file-uploads/rtdna/Research/2021_RTDNA- Syracuse_Research_-_News_Business_Trends_and_Threats_1_.pdf
Patterson, T. (2025). Can They Do Good and Still Do Well? Local TV Stations and Communities’ Information Needs. Retrieved from https://shorensteincenter.org/can-good-still-well-local-tv-stations-communities- information-needs/
Peterson, E. (2017). Paper cuts: how reporting resources affect political news coverage.
Program in Quantitative Social Science. Dartmouth College. Hanover, NH.
Pew Research Center. (2019). For local news, Americans embrace digital but still want strong community connection. Retrieved from https://www.journalism.org/2019/03/26/nearly-as-many-americans-prefer-to-get- their-local-news-online-as-prefer-the-tv-set/
Shearer, E., Matsa, K. E., Lipka, M., Eddy, K., & Forman-Katz, N. (2024). Americans’ Changing Relationship With Local News. Retrieved from https://www.pewresearch.org/journalism/2024/05/07/americans-changing- relationship-with-local-news
Snyder, J., & Stromberg, D. (2010). Press Coverage and Political Accountability | Journal of Political Economy: Vol 118, No 2. 355-408. Retrieved from https://www.journals.uchicago.edu/doi/abs/10.1086/652903
Soto, A. J., Mohammad, A., Albert, A., Islam, A., Milios, E., Doyle, M., . . . Ferreira de Oliveira, M. C. (2015). Similarity-based support for text reuse in technical writing. Paper presented at the Proceedings of the 2015 ACM Symposium on Document Engineering.
St. Aubin, C., & Naseer, S. (2023). Trends and Facts on Local News. Retrieved from https://www.pewresearch.org/journalism/fact-sheet/local-tv- news/?tabItem=454bdd61-e709-414f-9037-5fd1b4b25ec6
Stilson, J. (2024). Sook Sees ‘Huge Opportunity’ For Nexstar In Relaxed Ownership Cap. Retrieved from https://tvnewscheck.com/business/article/sook-sees-huge- opportunity-for-nexstar-in-relaxed-ownership-cap/
Tameez, H. (2023). OpenAI will give local news millions to experiment with AI. Retrieved from https://www.niemanlab.org/2023/07/openai-will-give-local-news-millions-to- experiment-with-ai/
TVB. (2023). 2022 Voter Funnel Study. Retrieved from https://www.tvb.org/research- measurement-analytics/research/category-studies/political-category- studies/2022-voter-funnel-study/
TVB. (2025). 2025 Media and Conversations Study. Retrieved from https://www.tvb.org/research-measurement-analytics/research/2025-media-and- conversations-study/
Waldman, S. (2011). The information needs of communities. Retrieved from Washington, DC: https://transition.fcc.gov/osp/inc- report/The_Information_Needs_of_Communities.pdf
Wenger, D., & Papper, B. (2018). Local TV News and the New Media Landscape. Retrieved from Miami, FL: https://knightfoundation.org/reports/local-tv-news-and-the-new- media-landscape
Weprin, A. (2024). Trump May Usher In Massive TV Station Land Grab. Retrieved from https://www.hollywoodreporter.com/business/business-news/nexstar-sinclair- tegna-tv-station-trump-1236060274/
Zhao, R. (2020). Will artificial intelligence replace human news anchors on television? In China, it is already happening. Retrieved from https://www.digitalcenter.org/web- insights/artificial-intelligence-news-anchors/
| Station Group |
# Owned |
# Controlled Stations |
% Controlled Stations |
| Nexstar |
127 |
138 |
16.0 |
| Gray |
114 |
118 |
13.7 |
| Sinclair |
81 |
86 |
10.0 |
| Tegna |
57 |
60 |
7.0 |
| EW Scripps |
42 |
43 |
5.0 |
| Comcast/NBC |
30 |
30 |
3.5 |
| Hearst |
29 |
29 |
3.4 |
| Mission Broadcasting |
25 |
17 |
2.0 |
| Univision |
24 |
25 |
2.9 |
| CBS TV |
23 |
23 |
2.7 |
| Fox |
19 |
19 |
2.2 |
| News-Press & Gazette Co |
17 |
19 |
2.2 |
| Meredith Corp |
17 |
17 |
2.0 |
| Quincy Media |
15 |
16 |
1.9 |
| Entravision |
13 |
12 |
1.4 |
| Allen Media Broadcasting LLC |
13 |
14 |
1.6 |
| Cox Media Group |
12 |
11 |
1.3 |
| Cunningham Broadcasting |
11 |
3 |
0.3 |
| Terrier Media Holdings |
10 |
10 |
1.2 |
| Standard Media Group LLC |
8 |
8 |
0.9 |
| ABC/Disney |
8 |
8 |
0.9 |
| American Spirit Media LLC |
6 |
2 |
0.2 |
| Lockwood Broadcasting |
6 |
7 |
0.8 |
| Cowles Publishing Company |
6 |
7 |
0.8 |
| Morgan Murphy Media |
6 |
7 |
0.8 |
| Hubbard Broadcasting |
6 |
6 |
0.7 |
| Graham Media Group |
6 |
6 |
0.7 |
| Morris Multimedia |
6 |
6 |
0.7 |
| Bahakel Communications Limited |
5 |
6 |
0.7 |
| SagamoreHill Broadcasting |
4 |
1 |
0.1
|
| Mumblow, Stephen |
4 |
4 |
0.5 |
| Coastal Television Broadcasting LLC |
4 |
4 |
0.5 |
| Vision Alaska |
4 |
4 |
0.5 |
| Griin Communications |
4 |
4 |
0.5 |
| Vaughan Media LLC |
3 |
1 |
0.1 |
| Capitol Broadcasting |
3 |
3 |
0.3 |
| Sunbeam Television Corp |
3 |
3 |
0.3 |
| Marquee Broadcasting |
3 |
3 |
0.3 |
| Block Communications |
3 |
3 |
0.3 |
| Forum Communications |
3 |
3 |
0.3 |
| Weigel Broadcasting Company |
3 |
3 |
0.3 |
| Lilly Broadcasting |
2 |
1 |
0.1 |
| White Knight Broadcasting |
2 |
1 |
0.1 |
| CP Media LLC |
2 |
3 |
0.3 |
| Calif-Oregon Broadcasting |
2 |
2 |
0.2 |
| VistaWest Media LLC |
2 |
2 |
0.2 |
| Circle City Broadcasting |
2 |
2 |
0.2 |
| Draper Communications |
2 |
2 |
0.2 |
| SDR Rockfleet Holdings LLC |
2 |
2 |
0.2 |
| GOCOM Media of Illinois LLC |
2 |
2 |
0.2 |
| Heartland Media LLC |
2 |
2 |
0.2 |
| Liberty Broadband Corp |
2 |
2 |
0.2 |
| Manship Family |
2 |
2 |
0.2 |
| Sarkes Tarzian |
2 |
2 |
0.2 |
| Thunder Bay Broadcasting |
2 |
2 |
0.2 |
| Galloway, Sheldon |
1 |
0 |
0.0 |
| Granite Broadcasting Corp |
1 |
0 |
0.0 |
| MPS Media |
1 |
0 |
0.0 |
| Maxair Media LLC |
1 |
0 |
0.0 |
| Terrier Media Holdings Inc |
1 |
0 |
0.0 |
| Glendive Broadcasting Corp |
1 |
0 |
0.0 |
| Horseshoe Curve Communications LLC |
1 |
0 |
0.0 |
| Montclair Communications |
1 |
0 |
0.0 |
| Woods Communications Corporation |
1 |
0 |
0.0 |
| SJL Broadcast Management Corp |
1 |
2 |
0.2 |
| Howard Stirk Holdings LLC |
1 |
2 |
0.2 |
| Waitt Broadcasting |
1 |
2 |
0.2 |
| Waterman Broadcasting Corp |
1 |
2 |
0.2 |
| 54 Broadcasting |
1 |
1 |
0.1 |
| Bangor Communications |
1 |
1 |
0.1 |
| Cadillac Telecasting Co |
1 |
1 |
0.1 |
| Heritage Broadcasting Group |
1 |
1 |
0.1 |
| Homan Communications |
1 |
1 |
0.1 |
| SagamoreHill Midwest LLC |
1 |
1 |
0.1 |
| SagamoreHill of Columbus GA LLC |
1 |
1 |
0.1 |
| Big Horn Television LLC |
1 |
1 |
0.1 |
| Community News Media LLC |
1 |
1 |
0.1 |
| Estrella Media |
1 |
1 |
0.1 |
| KQDS Acquistion Corp |
1 |
1 |
0.1 |
| Palm Broadcasting Co LP |
1 |
1 |
0.1 |
| Ramar Communications |
1 |
1 |
0.1 |
| Red River Broadcast Company LLC |
1 |
1 |
0.1 |
| Roberts Media LLC |
1 |
1 |
0.1 |
| Second Generation Television |
1 |
1 |
0.1 |
| America-CV Station Group |
1 |
1 |
0.1 |
| Berkshire Hathaway Inc |
1 |
1 |
0.1 |
| Bonneville International Corporation |
1 |
1 |
0.1 |
| CNZ Communications LLC |
1 |
1 |
0.1 |
| Ft Myers Broadcasting |
1 |
1 |
0.1 |
| KTBS LLC |
1 |
1 |
0.1 |
| Korean American TV Broadcasting Corp |
1 |
1 |
0.1 |
| Lincoln Broadcasting |
1 |
1 |
0.1 |
| Maranatha Broadcasting |
1 |
1 |
0.1 |
| Meruelo Media Holdings LLC |
1 |
1 |
0.1 |
| Mountain Broadcasting |
1 |
1 |
0.1 |
| NPM Inc |
1 |
1 |
0.1 |
| Paxton Media Group LLC |
1 |
1 |
0.1 |
| Rapid Broadcasting Company |
1 |
1 |
0.1 |
| Seal Rock Broadcasters LLC |
1 |
1 |
0.1 |
| Southeastern Ohio Television System |
1 |
1 |
0.1 |
| Telephone & Data Systems |
1 |
1 |
0.1 |
| Texas Television |
1 |
1 |
0.1 |
| Thomas Broadcasting Company |
1 |
1 |
0.1 |
| Tyler Media Group |
1 |
1 |
0.1 |
| University of Missouri |
1 |
1 |
0.1 |
How to Cite this web report
Yanich, Danilo & Bagozzi, Benjamin E. Reusing the News: Duplicating Local TV Content, University of Delaware, 2025.