Skip to Content View Previous Reports

Content Analysis

Content Analysis

Look closely at the content of cable news and it becomes clear that its appeal is its ubiquity and convenience; the medium does not come close to delivery on the potential of its depth or breadth.

Those are some of the key findings of the Project’s new two-pronged approach to examining cable news. As we did last year, we first studied five sample days of each of the three cable news networks for sixteen hours each day, or 240 hours of programming. That provided us with a sense of the types of stories and the level of repetitiveness that appeared over the course of a cable news day.

In addition, this year we selected three different types of programs from each of the three channels to study for twenty days to see how their choice and treatment of topics compared with other media studied the same days. The programs studied included one prime-time talk show, one daytime news show and the closest each news channel comes to producing a traditional evening newscast. That added up to 180 more hours of programming involving another 4,551 stories.

Repetition vs. Updating

Although cable news purports to provide continuously updated coverage of breaking news, the idea that it is “following” stories and adding new information through the course of the day is in the main an illusion.

In the course of sixteen hours of viewing starting at 7 a.m. for five separate days, most of the stories on cable news (67%) are the same matter turned to repeatedly, and only 10% add meaningful updates with substantive new information.

In other words, 60% of all stories aired on cable through the day are simple repetition of the same information. Just one in three stories in the course of a cable day is new, or something not aired earlier.

Those figures are nearly identical to what we found last year, when 68% of the stories were repetitious, just 5% contained any substantive updates, and 27% were completely new.

What does that mean? With hours of air time and numerous correspondents, resources are devoted much less to gathering new information, or going deeper with background reporting, than to being live and appearing to be on top of three or four big stories of the day.

Repetition on Cable News
Percent of all stories from 7A.M.-11P.M.

Exact Repeat
12%
Repeat: No New Substance
35
Repeat: New Angle
11
Repeat: New Substance
10
New Story
33
Totals may not equal 100 due to rounding.

Breadth of Topic

The consequence is a notably limited breadth of reporting. In all, the three cable programs we followed over twenty days tended to cover a narrower range of topics even than network evening newscasts — and far less than online or print. Cable news spends a smaller percentage of its time than does network evening news covering government and, perhaps even more notably, roughly half as much time covering the broad range of domestic issues, from the environment, to transportation, health care, social security, welfare, education, economics, technology, science and more.

In contrast, lifestyle, entertainment and celebrity — topics virtually nonexistent on nightly newscasts or the front pages of newspapers — are the largest topic group on cable news. And that holds true even though the amounts vary across the range of program types. For instance, collectively, science, technology, and business made up just 2% of the time studied over twenty days, and the range of domestic issues, from education to the environment to health care, made up 11%. Celebrity, lifestyle and entertainment made up nearly a quarter of the time (23%).

Topics on Cable and Network News
Percent of all time

Cable
Network Evening
Network Morning
Government
17%
29%
25%
Defense/Military
7
1
0
Foreign Affairs
9
14
8
Elections
14
11
8
Domestic Affairs
11
20
15
Business
1
4
1
Crime
3
1
5
Science/Technology
1
4
3
Celebrity
14
2
4
Lifestyle
9
4
7
Accidents/Disasters
2
4
3
Other
12
6
21

Totals may not equal 100 due to rounding.

Live Reporting Lives On

The second major feature of cable news is that it is dominated by the culture of live, extemporaneous journalism.

Over the three programs studied for twenty days, 52% of time was spent in live interviews (usually by anchors) or reporters in live standups.

The medium, as noted last year, “has all but abandoned what was once the primary element of television news, the written and edited story.”

Less than half as much time, 24%, on the cable programs studied is made up of correspondent packages. Compare that to network nightly newscasts, in which 86% of time is such packages, or even morning news or PBS, where a third of time is correspondents telling stories.

Another 17% of time is devoted to anchors reading the teleprompter, so-called tell stories, either with video or without. And, in a feature that is not usually found on other TV programming, 6% of the time covered live events such as press conferences. (1% was spent on banter between stories.)

Story Origination on Cable News
Percent of all time

Packages
Staff Package
24%
Staff Live
52
Anchor Voice Over/Tell Story
17
Live Events
6
Banter
1
External Outlets
1
Totals may not equal 100 due to rounding.

The figures for those three programs studied in depth, moreover, closely mirror what viewers would see on the cable networks generally. In the course of a sixteen-hour day, 46% of the time on cable is spent in what its producers can describe as live — either live interviews, usually by anchors, or reporters talking live to the camera. Another 5% is live events such as press conferences aired as they happen (usually without a reporter on camera in any form).

Only 26% of the entire cable day comprises correspondent packages. Anchors reading the teleprompter, so-called tell stories or headline news, with or without video, account for another 20%.

Promotions and small talk between staff members take up the remaining 5% of the day, or 48 minutes of air time.

The approach has all the virtues of capturing events as they are happening. It is also cheaper, and helps create the impression that things are up to date.

But the amount of updating, as we noted, is minimal, and the emphasis on live cable news has resulted in walking away from the capacity to review, verify, edit, choose words carefully and match those words to pictures.

Audiences are even less likely to find verified, edited journalism at certain times of the day. Those watching talk shows such as Larry King or Bill O’Reilly will see barely any taped packages. Those watching the closest thing that cable has to a signature evening newscast (Brit Hume on Fox, Aaron Brown on CNN or Keith Olberman on MSNBC) are more likely to see taped packages (42% of all time). Still, that is only half as likely as on the broadcast evening news on ABC, CBS or NBC, where 86% of all time is edited packages.

Thinner Reporting on Cable

In part because of the dependence on being live — and the illusion that creates of being new — cable news is also more thinly reported than most other kinds of national TV news.

Over all, cable news stories have fewer sources than those on broadcast, reveal less about those sources, and, if the story involves a dispute, contain fewer conflicting points of view than in broadcast TV.

To pin this down, the study went deeper this year than last in examining the depth or thoroughness of reporting. This analysis was done in the three key parts of the day studied for 20 days — daytime, prime-time talk and the key evening newscast — to get a range of different styles of programming, and to match the same days studied in other media. Specifically, we studied:

Depth of Sourcing

In general, cable news was less likely than other media to contain multiple sources with enough description attached — their identity, their level of knowledge about the events being described and any potential biases — to enable audiences to judge what they were saying.

Only a quarter of cable stories studied (26%) contained even two or more such sources. That compares with 50% of network evening news stories, 81% of stories on newspaper front pages and 78% of online news stories. Even network morning shows, with their penchant for long one-person interviews, tended to have significantly more stories, 39%, with at least two fully transparent sources.2

Most cable stories (74%) had no source that audiences could fully identify, or only one.

The dependence on live programming is one reason cable reporting is thinner and at the same time less transparent. The live reporting on cable is even more thinly sourced than cable news as a whole. Most of the live reports, 60%, were based on only a single source that audiences could fully identify. The taped edited packages on cable were four times as likely to contain four or more fully identified sources as the live reports, and nearly twice as likely to contain two or three (see chart). But even the taped, edited packages on cable contained fewer fully transparent sources than packages on commercial broadcast newscasts or on PBS, despite cable’s advantage of having more time for the news.

Source Transparency on Cable, by Story Type
Percent of all stories

Packages
Live Reports/
Interviews
Anchor Voice Over
Anchor Reads
Live Events
None
12%
11%
78%
74%
20%
1
23
60
18
16
63
2-3
45
25
3
9
11
4
20
5
1
1
6
Totals may not equal 100 due to rounding.

Breadth of Viewpoints

The reporting on cable news is also more one-sided than that in other media studied.

Over all, only a quarter of cable stories that involved controversy contained anything more than a passing reference to a second point of view. That was much less balanced than all the over-the-air broadcast news programs studied. Indeed, stories on morning news, PBS evening news and those on newspaper front pages were more than three times as likely to contain a mix of views, and commercial evening newscasts just under that.

Range of Viewpoints on Cable and Network News
Percent of all stories

Cable
Network Evening
Network Morning
Mix
27%
72%
86%
Mostly One View
21
8
2
Only One View
52
20
11
Totals may not equal 100 due to rounding.

Certain kinds of storytelling on cable tended to be more balanced than others, and again live reporting was at the bottom of that scale. More than three-quarters of interviews and reporter standups (78%) told only one side, or mostly one side, of controversial stories. That meant only 22% of live reported stories offered a balance of at least two viewpoints.

Taped packages on cable fared better, but not dramatically. Just over a third of the taped packages studied (38%) offered at least two points of view, which still meant that 62% were mostly one-sided. Indeed, the taped edited packages on cable do not stack up against those on network news in this regard. On the Big Three commercial nightly newscasts, 75% of the taped packages contained multiple viewpoints. So the style of storytelling does not entirely explain the one-sidedness of cable.

Range of Viewpoints on Cable News, by Story Type
Percent of all stories

Packages
Staff Live
Anchor Voice Over
Anchor Reads
Live Events
Mix
38%
22%
30%
32%
9%
Mostly One View
29
21
10
7
4
Only One View
34
58
60
61
88
Totals may not equal 100 due to rounding.

Live reports also differ from taped packages in ways some people might argue are advantages. For instance, while reporter standups and live interviews tended not to cite multiple sources, they also tended to avoid citing anonymous sources, perhaps because they often had just one source in all — the interviewee. Just 5% of live reports on cable contained anonymous sourcing, compared with 20% of packaged pieces.3

Interestingly, correspondents and anchors on live and unscripted stories also seem less likely to inject their own opinions in their reports. Just over a third of live cable reports, 34%, contained journalist opinion, versus 43% of packaged pieces.

One possible explanation is that reporters and anchors who are live may adopt a stenographic frame of mind, trying to simply recall and recite what they have been told. That would help explain both their tendency toward one-sidedness and their avoidance of giving opinions.

Journalist Opinion on Cable News, Select Story Types
Percent of all stories

Packages
Staff Live
Total
No Opinion
57%
66%
71%
Opinion
43
34
28
Totals may not equal 100 due to rounding.

One area where there is little difference between live and packaged is how correspondents frame their reporting. Packages tend to be a little more conflict-oriented while live reports do a little more reality check pieces (i.e. is this really true? What does this really mean?) and telling of a good tale. Generally, though, correspondents gravitate to the same kinds of frames.

Story Frames on Cable, Select Story Types
Percent of all stories

Packages
Staff Live
Total
Conflict
42%
31%
23%
Consensus
4
5
3
Winners/Losers
12
7
5
Problems to Solve
8
5
4
Good Yarn
10
13
10
Reality Check
2
5
2
Underlying Principles
2
4
2
Other
5
5
4
No Frame
15
25
45
Totals may not equal 100 due to rounding.

Are interview-based programs necessarily less able to offer a broad range of views and deep sourcing? Perhaps not. One interview-based program that seems to do a good job of this in the PBS NewsHour, which often combines packages and live discussion. (See Network/Content Analysis)
Differences Among Cable Channels

Our content analysis also shows measurable differences in what each of the cable networks puts on the air. This study made no attempt to identify bias, or whether one network tilted to the Democrats or Republicans. Some more basic distinctions, however, were evident.

Fox was measurably more one-sided than the other networks, and Fox journalists were more opinionated on the air. The news channel was also decidedly more positive in its coverage of the war in Iraq, while the others were largely neutral. At the same time, the story segments on the Fox programs studied did have more sources and shared more about them with audiences.

CNN tended to air more points of view in its stories than others, and its reporters rarely offered their own opinions, but the news channel’s stories were noticeably thinner in the number of sources and the information shared about them.

MSNBC consistently fell between its two rivals on most indices.

In the degree to which journalists are allowed to offer their own opinions, Fox stands out. Across the programs studied, nearly seven out of ten stories (68%) included personal opinions from Fox’s reporters — the highest of any outlet studied by far.

Just 4% of CNN segments included journalistic opinion, and 27% on MSNBC.

Fox journalists were even more prone to offer their own opinions in the channel’s coverage of the war in Iraq. There 73% of the stories included such personal judgments. On CNN the figure was 2%, and on MSNBC, 29%.

The same was true in coverage of the Presidential election, where 82% of Fox stories included journalist opinions, compared to 7% on CNN and 27% on MSNBC.

Those findings seem to challenge Fox’s promotional marketing, particularly its slogan, “We Report. You Decide.”

Some observers might argue that opinions clearly offered as such are more honest than a slant subtly embedded in the sound bites selected or questions asked. But that was not the case here. Given the live formats on cable, the opinions of reporters and anchors are often embedded in questions or thrown in as asides. Only occasionally were they labeled as commentary.

Journalist Opinion in Iraq War Coverage, Cable News
Percent of Iraq War stories

CNN
Fox
MSNBC
Total
No Opinion
98%
27%
71%
70%
Opinion
2
73
29
30
Totals may not equal 100 due to rounding.

Tone of Coverage

The study this year also tried to assess the tone of coverage.4 When it came to the war, Fox again looked different from the others by being distinctly more positive than negative. Fully 38% of Fox segments were overwhelmingly positive in tone, more than double the 14% of segments that were negative. Still, stories were as likely to be neutral as positive (39%) and another 9% were multi-subject stories for which tone did not apply.

On CNN, in contrast, 41% of stories were neutral in tone on the 20 days studied, and positive and negative stories were almost equally likely — 20% positive, 23% negative. Some 15% were multi-faceted and not coded for tone.

MSNBC’s stories about the war were most likely to include several issues or subjects, so that no one area could be coded for tone. Fully four in ten stories were of this nature. Otherwise, the network’s coverage, like CNN’s, was more neutral (28%) with positive and negative stories almost equally prevalent, (16% positive and 17% negative).

Tone of Iraq War Coverage on Cable News
Percent of Iraq War stories

CNN
Fox
MSNBC
Total
Positive
20%
38%
16%
24%
Neutral
41
39
28
36
Negative
23
14
17
19
Multi-Subject
15
9
40
21
Totals may not equal 100 due to rounding.

When it came to election coverage, the majority of stories on every network had no overwhelming tone. Here MSNBC stood out as being twice as likely to air candidate and issue stories with a positive tone as with a negative tone. CNN’s coverage, on the other hand, was more likely to be negative. Fox was divided equally among positive and negative stories.

Tone of Election Coverage on Cable News
Percent of election stories

CNN
Fox
MSNBC
Total
Positive
10%
16%
17%
15%
Neutral
62
56
32
47
Negative
17
17
8
13
Multi-Subject
11
12
42
25
Totals may not equal 100 due to rounding.

Only weeks after being installed as CNN’s president, Jonathan Klein proclaimed an end to the shout fests that have come to characterize cable news, canceling the network’s archetypical Crossfire program and declining to renew the contract of the conservative talker Tucker Carlson. “We always want to be provocative,” Klein said. “But there is a numbness that has set in among those head-butting festivals. I’m convinced that the political brainiacs we have at CNN can come up with a better way to engage the audience.”

In place of shouting, Klein said, he wanted to return to “roll-up-your-sleeves storytelling.”

“CNN is a different animal,” Klein told the New York Times. “We report the news. Fox talks about the news. They’re very good at what they do and we’re very good at what we do.”5

Is there evidence that CNN is more fact-oriented, more neutral and more tied to storytelling than rivals Fox or MSNBC in 2004? Where does each station fall heading as the new year unfolds?

CNN, according to the data, does indeed seem to offer more neutral reporting. Its adherence to storytelling, though, seems to be more of a mixed bag. Its NewsNight with Aaron Brown is heavy on such pieces, but its noontime programming spends less time on packaged pieces than Fox or MSNBC.

MSNBC fits somewhere in the middle on most of these measures, perhaps waiting to see which approach bears most fruit.

Three Distinct Types of Programs

Last year, the study found that the cable day broke down into four distinct parts of the day: the traditional morning show, daytime, early evening and prime time. Each had its own personality, with the three networks remarkably similar within each time frame. This year, to look more closely at those dayparts, we examined an hour of daytime, a prime-time talk show and the closest thing that each of the network offers to a prime-time signature newscast, all in a 20-day period.6

Prime-Time Talk Shows

The highest-rated program on every network is a prime-time talk show, and we examined each of them: Larry King on CNN, Bill O’Reilly on Fox and Chris Matthews on MSNBC.

The three shows are built around interviews, which take up 81% of their time, but they are not identical. King leans almost entirely on interviews — 95% of all his time. The O’Reilly Factor relies on them heavily as well (79%), but 20% of the program’s time is made up of the host reading news items and commentaries of his own.

MSNBC’s Hardball with Chris Matthews has evolved into something different. A quarter of its airtime (26%) is packaged pieces. Interviews with guests and MSNBC reporters make up 67% of the time. The remaining 8% is briefs and voice-overs. The channel appears to be trying to morph Hardball into something that is partially a news program.

The talk shows build their appeal partly around their hosts, of course, and partly around the celebrity of their guests rather than issues or events. As a group, these shows dedicated most of their time to three main topic areas in 2004: the elections (17%), government (16%), and celebrity/entertainment (16%). Lifestyle stories accounted for another 11% of the airtime and domestic affairs 14%.

Yet in choice of topics, the three programs also had different characters. Larry King devoted close to half (45%) of his time to entertainment and lifestyle topics, more than twice the figure for O’Reilly (21%) and three times as much as for Matthews (13%). Matthews, a former congressional press aide, spent more than half of his time on government and election topics, (the next most popular topics on CNN and Fox). O’Reilly’s program was more of a mix.

Beyond topic, the most striking difference among the three shows is in the presence of the host’s opinion. Nearly every story on Fox’s O’Reilly Factor (97%) contained O’Reilly’s opinions, even his quick news briefs. CNN’s Larry King was nearly the reverse, with only 2% of segments including his opinions. And despite to his reputation for dominating the guests, Chris Matthews on Hardball offered his opinion just 24% of the time.

Topics on Cable News Programs
Percent of all time

Total
Daytime
Newscast
Interview
Government
17%
18%
18%
16%
Defense/Military
7
6
6
9
Foreign Affairs
9
10
13
4
Elections
14
8
18
17
Domestic Affairs
11
12
10
11
Business
1
2
1
1
Crime
3
5
2
2
Science/Technology
1
1
1
7
Celebrity
14
15
10
16
Lifestyle
9
11
6
11
Accidents/Disasters
2
3
2
*
Other
12
9
13
6
Totals may not equal 100 due to rounding.

Cable’s Version of the Evening News

None of the cable channels airs a traditional evening newscast, but each has programs that come closer than others: Special Report with Brit Hume on Fox, NewsNight with Aaron Brown on CNN, and Countdown with Keith Olbermann on MSNBC.

The shows do more traditional storytelling than other cable programs, but produced tape packages are still only half as prevalent as on commercial network news — 42% of all time versus 86%. In format, these news programs are closer to network morning news, or PBS, where packages make up a third of time.

In the topics they cover, the cable shows also differ from their broadcast counterparts. They cover government less (18% vs. 29% of time on network), and the broad range of domestic issues half as much (10% versus 20%). Meanwhile celebrity and lifestyle, virtually non-existent on broadcast nightly newscasts, account for 16% of the time on their cable counterparts. The only topic that gets similar amounts of time on both cable and broadcast network evening newscasts is foreign affairs.

In depth of sourcing, the news round-up programs do better than their cable siblings, but again fall short of levels found on network evening news or in print. That pattern also holds true for the mix of viewpoints offered.

Source Transparency on Cable versus Network News
Percent of all stories

Cable Daytime
Cable Newscast
Cable Talk Shows
Network Evening
None
56%
38%
11%
37%
1
28
26
60
14
2-3
14
23
26
32
4
3
13
3
18
Totals may not equal 100 due to rounding.

Range of Viewpoints, Cable versus Network News
Percent of applicable stories

Cable Daytime
Cable Newscast
Cable Talk Shows
Network Evening
Mix of Opinions
18%
39%
26%
72%
Mostly One Opinion
24
28
13
8
All One Opinion
59
33
61
20
Totals may not equal 100 due to rounding.

Are there differences between the networks here? Aaron Brown’s program on CNN leans the most on taped edited packages — double Keith Olberman on MSNBC and more than Brit Hume on Fox. Olbermann favors summaries of the news (more than triple that of CNN or FOX). Hume is a mix of taped packages followed by interviews. The Hume program is also Washington-centric — 60% of it concerns government, the military and politics).

Daytime

The bulk of the cable news networks’ time, between 5 and 7 hours a day, is made up of programs that might be called dayside. These shows, between 9 a.m. and roughly 3 p.m. depending on the channel, track the news of the day as it is happening.

The dayside programs offer a broader mix of storytelling formats than anything else we studied on cable. Edited packages take up 20% of the time, live interviews and standups 36%, and anchors’ reads, sometimes with pictures, another 22%. Shown during the workday, these programs are more likely than others to carry events live (18% of time).

Yet like their evening counterparts, these programs are conspicuously limited in their range of topics. Entertainment and lifestyle stories were given the most attention — a quarter of all airtime, (26%) nearly exactly the same as on prime-time cable talk shows. Interestingly, the daytime programs studied devoted less time to the elections (8%) than the other cable programs.

Sourcing on these daytime news programs was measurably thin. More than half, 56%, of all stories had not even a single fully identified source. Another 28% had just one. A mere 3% of all stories contained four or more fully transparent sources.

Journalist Opinion on Cable News Programs
Percent of applicable stories

Daytime
Newscast
Talk Show
No Opinion
68%
74%
73%
Opinion
31
26
27
Totals may not equal 100 due to rounding.

Despite what Klein of CNN suggested, the daytime programs we studied were even less focused on the storytelling he was referring to than the rival networks.

At noon, CNN Live Today was the least devoted to packaged pieces (roughly half as much time as Fox News Live and a third as much as MSNBC Live). It spent more time instead on quick anchor reads. And lifestyle stories made up more than 20% of all airtime.

Fox News, on the other hand, tended to spend more time covering live events, and as a result offered more coverage about the government than either of the other stations.

Summary

For the second straight year, content analysis raises substantial questions about the nature of reporting on cable news. The time required to continuously be on the air seems to take a heavy toll on the nature of the journalism presented. While there are differences between channels identified this year in coding of the thoroughness of the reporting, the sector generally falls behind those of other media studied.

It appears that the appeal of cable is its convenience. It is there when you need it, and in a nation of multi-taskers, it can be on as a kind of background, something we can turn to in moments of curiosity.

The problems exposed in the content analysis may begin to seem more troubling to viewers when the Internet in the next year or two begins to meaningfully add searchable video. At that point, the Web will begin to present television on demand, when you want it, and in a searchable form.

Then the second disadvantage of cable as an on-demand medium will become more important: the fact that one has to sit through “the wheel” of whatever is on before a subject of choice might appear.

The Internet will offer the advantages of carefully produced packages, with the convenience of having it there when you want it.

The question will be how much hold the ease of television has on viewers — it comes at you without your having to so much as click a mouse — combined with the impression of its being up to the minute because it is “live.”

Footnotes

1. For a more detailed explanation of depth of sourcing, please see the Methodology.

2. Only 6% of cable stories contained four such sources, also far less than other media.

3. We also saw in broadcast network news that the more transparent sources a story contains, the more likely it is to also include an anonymous source. This may also be the case here, as packaged pieces contain more transparent sourcing, they also contain more anonymous sources as well.

4. Some observers are skeptical about the concept of assessing tone in content analysis. Taking those doubts into consideration, the Project assesses tone differently from many other analysts. Rather than simply tallying up statements as positive, negative or neutral, the Project analyzes the assertions in the story and then measures whether they tilt the entire story to be clearly positive or negative. For a story to have a distinct tone one way or the other, the negative or positive assertions must outweigh all others by at least two to one. What’s more, we studied tone only in stories in which a single issue or newsmaker was the clear focus of at least half the story.

5. Bill Carter, “CNN Will Cancel ‘Crossfire’ and Cut Ties to Commentator,” The New York Times, January 6, 2005.

6. We did not include a morning program for the cable networks for two reasons. First, the morning shows air so early on the West Coast that their ratings are miniscule – unlike network TV, in which the morning shows are delayed and run from 7 a.m. to 9 a.m. everywhere. Second, one of the three networks, MSNBC, airs only a static camera of Don Imus’s syndicated radio program.

Daytime programming is pretty much the same format from 11 a.m. to 3 or 4 p.m. depending on the network. We choose to examine the hour from 11 a.m. to 12 p.m.

After that, the cable channels begin counter-programming each other – airing similar programs at different times, to try to capture the greatest audience. For instance, the news program that is considered by most to be the closest thing to a network newscast airs at 6 p.m. on Fox, 8 p.m. on MSNBC, and 10 p.m. on CNN. Prime-time talk shows are similarly countered. For this reason, we chose the next two groupings not according to the time they aired but according to the type of program — newscasts and prime-time talk shows.

Click here to view content data tables.