Skip to Content View Previous Reports

Content Analysis

Content Analysis

What does the typical local television newscast look like? For some chapters of this report, the Project did a special content analysis of media sectors in 2003. No such content study was done for local television news. From 1998 through 2002, however, the Project undertook an exhaustive examination of the content of local news in 15 to 20 markets a year – randomly selected with controls for station size and geographic diversity. In all, 154 stations were studied, some multiple times. Based on that research, the largest single study of local news we know of, the Project has a fairly detailed sense of what Americans generally get from local television news.1

This detailed analysis suggests that local television news is, indeed, very local. It does a good job of covering many everyday events in most communities, especially incidents from the police scanner and events on the news daybook. Local television news is, also, not the same everywhere.

Yet the data also reveal that the medium is dominated by the ethos of “live, local and late-breaking” coverage, particularly of crime. Many of the stories are formulaic, reactive and, above all, short.

Here are some findings:

What Gets Covered

The Project’s study found a remarkable consistency in terms of what got covered, a surprising range in terms of quality, an alarming tendency toward one-sidedness and a steady disappearance of enterprise.

In every year of the study, the No. 1 topic on local television news was crime. Over the five years, 24 percent of all stories were about crime – the number ranging from a high of 26 percent in 2002 to a low of 19 percent in 2000. (Some stations clearly go the other way. In 2002, the last year of the study, there were stations where crime represented as little as 5 percent of stories. But these were unusual.)

If crime was the No. 1 topic on local news by more than 2 to 1, what came next? Over the five years, stories about accidents, bizarre events, fires and catastrophes accounted for 12 percent. Taken together, crime, fire accidents and disasters made up 36 percent of all stories.

The next most popular topics were human-interest stories and politics, which each accounted for 10 percent of stories. This was followed by social issues – a category that includes a wide swath of topics such as education and transportation – at 8 percent, and then by business and economic news at 7 percent, although most of these stories were brief updates on Wall Street rather than larger pieces on economic issues. (The study did not attempt to track weather and sports coverage.)

Thus the notion that it has to bleed to lead in local television news is an exaggeration, but it is grounded in some reality. We looked just at lead stories on the 2,400 newscasts studied to test this idea. We found that 39 percent were crime stories, while 13 percent were about disasters or severe weather and 9 percent were about a fairly routine fire or accident, for a total of 61 percent. Local TV news is not dominated by these topics, but it does tend somewhat to lead with them.
Topic Coverage of Local TV News Stories

Topic Percent of Stories
Crime/trials 24%
Accidents/bizarre events/disasters 12
Politics/government 10
Human interest 10
Social issues 8
Business/economy 7
Culture/civilization 7
Health/consumer 6
Miscellaneous 6
Foreign affairs/defense 5
Science/technology 4

Source: PEJ Local TV News Project.
Weather and sports coverage not counted towards total. Totals may not equal 100 due to rounding.

Story Focus

The study also looked at what aspect of an event a story focused on regardless of topic. To what extent did the story try to extract an underlying idea or trend? To what extent was it focused on just the incident itself? Was it that there was crash at North and Main Streets? Or did the story note that this was the fifth crash in a month at that corner? Or did the story go so far as to explain why the problem corner had not been fixed with a traffic light?2

Here, the reactive and somewhat superficial nature of so many stories becomes apparent. In all, 41 percent of stories focused on everyday incidents, such as relatively common car accidents or crimes. Just 15 percent dealt with ideas by finding the trend or larger problem suggested by whatever the incident was. Eleven percent concerned unusual events, things that were not typical or happened every day. Only 8 percent focused on the health of local institutions.

Story Focus of Local TV News Stories

Story Focus Percent of Stories
Everyday incidents/crimes 41%
Human interest/pop culture/scandal 15
Ideas, issues, or trends 15
Unusual/monumental events 11
Local institutions 8
Political strategy 6
Breaking events 2
Public malfeasance 1
Other 1

Source: PEJ Local TV News Project
Totals may not equal 100 due to rounding.

Actors in the News

Who do we see, and not see, on local television news?

Here again, the orientation toward live local and late-breaking news, especially of crime, has an impact. Suspects, lawyers, police officers and victims of crime made up 26 percent of all the people seen in the local news study over five years. Firefighters and emergency medical personnel add 3 percent more.

Who was not seen on local news also stands out. Seniors, for instance, were an asterisk, not even 1 percent. The poor and immigrants also did not even register a single percentage point as on-camera subjects in stories.

Government officials of all sorts, from the president to city council members, made up 14 percent, the second-largest group – after people involved in crime.

Main Subjects of Local TV News Stories

Main Subject Percent of Stories
People involved in crime (e.g., victims, criminals, law enforcement, judges) 26%
Elected officials (e.g., Congress, mayors, school boards) 14
Business/corporations 8
Natural events (floods, severe weather, etc.) 5
Celebrities and athletes 5
Participant in unique events 5
Person in the street 4
Other 33

Source: PEJ Local TV News Project
Totals may not equal 100 due to rounding.

Story Length

Another characteristic of local television news is that the stories are short. In the study:

This brevity may help stations cover a lot of news, but it tends to reinforce the sense of superficiality and lack of real connection to people. Faced with competition from all-news cable and the need to “grab the eyeballs” of people who have been watching lead-in entertainment programming, local newsrooms often appear to opt for immediacy over storytelling. Stations tend to emphasize this approach somewhat more in late news, following prime time, than in early news, at 5:30 or 6 p.m. When correlated with viewership trends, the findings suggest that airing more short stories may be unwise for stations trying to build viewer loyalty. In each year of the study, stations with better measures of commercial success (defined as their ability to attract or retain audience) tended to do fewer very short stories and more long stories. The problem is, of course, finding the right mix. A station that did only long stories, in a 22-minute newscast that includes sports and weather, would not cover much of the community.

Local vs. National

Local television news is also predominantly local. Three quarters of the stories (76 percent) were local and the balance were national or international.

One might think that this is, in part, because the charge of local television news is to be local. And it is. But that mandate is often tossed aside for reasons that belie local values. For instance, wildfires in California, which are highly visual, tended to lead local newscasts in Washington, D.C., even though they were far away. But other major national stories, such as economic news or politics, might never lead local newscasts.

Enterprise in Local News

The other characteristic of local television news is that it tends to react to events rather than seek out the news, increasingly so. The study also coded every story studied according to a hierarchy of enterprise. At the top were those stories labeled by the station as investigative, followed by non-investigative series, followed by tough on-camera interviews. At the bottom were video press releases put on the air. Sending an on-camera reporter to an event was scored higher on the enterprise ladder than just sending a camera. The data showed a growing reliance on stories that did not involve having a correspondent cover a story.

In the first year of the PEJ study, for instance, 14 percent of stories were feeds from somewhere else. In 2002, the fifth year of the study, the use of feed material had risen to 23 percent. Similarly, in the first year of the study, 9 percent of stories were prearranged events, such as press conferences, in which the station sent a camera but had no on-camera correspondent. By the fifth year, that, too, had risen, to 29 percent of all stories.

Also:

Why is this the case? One reason, interviews in newsrooms suggest, is the sense that yellow police tape and flashing lights make for good television. But, to the contrary, studies correlating viewership trends to story lineup choices suggest that leading newscasts with such stories does not build viewership.

Enterprise in Local TV News Stories

Form of Enterprise Percent of Stories
Coverage of “daybook” events with reporter present 26%
Coverage of “daybook” events with no reporter on camera 22
Wire/feed stories and stories attributed to another news source 22
Coverage of spontaneous events 21
High-level enterprise (e.g., investigations, news series, and tough interviews) 7
Other 3

Source: PEJ Local TV News Project.
Totals may not equal 100 due to rounding.

One-Sidedness in Local News

The study also examined the balance of sourcing by first asking whether the story involved a controversy or dispute. “The school board approved a new bond issue today,” may be presented as fact. Whether or not all the citizens whose tax bills may be affected by the decision agree with the new bond is a matter that is probably in dispute. In stories involving such disputes, did the reporting involve a mix of viewpoints, tell mostly just one side, or exclusively one side?3

The study consistently found a distinct one-sidedness to controversial stories. Indeed, in stories that involved some controversy, 60 percent of the stories told mostly or only one side. Another 40 percent contained a clear mix of opinions. This was true every year of the study, and though there were differences each year, no pattern emerged to suggest that this was an issue of resources or people being increasingly pressed for time.

Balance of Viewpoints in Controversial Local TV News Stories

Story Viewpoint 1998 1999 2000 2001 2002 Total
All one opinion 43% 55% 46% 50% 43% 47%
Mostly one opinion 12 16 9 13 11 13
Mix of opinions 45 29 44 38 45 40

Source: PEJ Local TV News Project
Totals may not equal 100 due to rounding.

Is Local News the Same Everywhere?

Some critics have suggested that local news has been so homogenized that it is the same everywhere – that the news in one town looks just like the news in every other with no real connection to the local community. One reason, observers have argued, is that a handful of three or four television news consulting firms dominated the industry from the 1970s through the 1990s and impressed the same small set of conventional ideas on virtually every station in the country. As a consequence, local television news has no sense of place.

Is that critique fair?

The data in the PEJ content study suggest that the complaint about homogenization appears to be partly true but exaggerated. For instance, crime was the most popular topic almost everywhere (at 191 out of 242 stations studied). Still, that leaves more than 50 stations at which crime was not No. 1. (At some stations, as noted above, crime was as low as 5 percent of stories.) The next-most popular topic varied – from human interest at some stations to politics at others.

There were some other differences that stood out as well. The most obvious was the variation in the number of stories aired in a typical newscast. The average number is 14 stories, but there were stations which ran as few as 7 per night to, in one case, a station that typically aired 27 stories.

The size of a station’s market had an effect on some aspects of local television news: stations in the smallest markets were less likely to air sensational footage and more likely to cover social issues.

What may be more true is that the style – the look and feel of local news – rather than the substance, seems homogenized. Local anchors tend to look alike and dress similarly. Almost all stations have adopted the “family model” – two anchors, usually a man and a woman, plus an amusing weathercaster and a jock newscaster. Older women are hard to find on newscasts, though older men are not. The graphics and sets are similar. So is the emphasis on live, local and late breaking. At least one reporter package starts with a live lead in, even if the event is hours old, and the story unfolds according to format. And many gimmicks – “The News at Six starts now” intones the anchor breathlessly – are so common that they are clichés. But the charge that the news is the same on every station everywhere is, beyond the style, unfair.

Footnotes

1. The content analysis in this section is based on a database of 33,911 stories from the 154 stations studied. Two weeks of half-hour news programs were studied at each station: one sweeps week, and one non-sweeps week. The timeslot selected for study was the half-hour when the most people in the market were watching local news. For more information, see Project for Excellence in Journalism, “Local TV News Project,” available online: http://www.journalism.org/resources/research/reports/localTV/default.asp.

2. The focus of the story was scored on a scale: for example, issues of public malfeasance were considered more important than stories about celebrities.

3. The study counted how many sources were used in each story, and how many points of view were presented. Stories presented as undisputed (a fire, the weather) were noted separately. In all other stories that involved some element of controversy, the study counted how many points of view were included – just one, mostly one, or a mix of more than one point of view.

Click here to view content data tables.