In the early days after September 11, critics praised the press for returning to a style of coverage that stuck more to reporting facts than interpreting them and for a notable caution about conveying rumor and speculation.

"The word of the day is steady, steady," Dan Rather said out loud on CBS the day of the attacks. "We are going to try to separate the rumors from the facts."

To examine whether this impression was true, the study looked at every story pertaining to the war on terrorism during the nine days studied.

Each statement or assertion was noted for whether it was: a), a fact b), a piece of analysis that could be attributed to some kind of reporting or c), an opinion or speculation that was un-attributed to anything. Each paragraph was then categorized by which type of statement predominated.

In the early days of September, the coverage was strikingly straightforward. More than three quarters (75%) of all the coverage was factual—here is what happened—as opposed to analysis or opinion.

Opinion and speculation accounted for just 11% of the reportage.

Analysis made up 14%.

The coverage was also notably well documented. Nearly half of the coverage (45%) cited four or more sources. More than three quarters of all sources were named (76%).

The reporting was highly factual and well sourced in September across the more traditional news genres—evening news, morning and newspapers. Facts were not as dominate on the talk shows (54%), the prime time hours (52%).

By November, the coverage began to shift, becoming more analytical. Factual reporting dropped by 12 percentage points to 63%.

Analysis rose by half, to 21%.

The amount of punditry grew to 15%.

In December, the numbers remained close to the November levels.

The level of documentation also shifted with time. By December, the percentage of stories citing four or more sources had dropped from more than four in ten to just above a quarter (29%). The percentage of stories citing just one source had grown from 20% to 25%.

Why the change?

One reason may be that in the earliest days, as Americans were digging out of the shock and rubble, a premium was put on avoiding undue panic and speculation. As the situation stabilized, and the war moved overseas, the temptation to analyze and speculate naturally increased.

Another factor may be access. In September, the story was largely a domestic one—and in the media's backyard in Washington and New York. Eyewitnesses and people with unique stories to tell were easier to find. The unrelenting financial cutbacks, particularly in television, were less of a factor when the events were occurring in the media's hometown.

As the war moved abroad, the Pentagon made access to soldiers and the battlefield more difficult than it has ever been. Web sites with previously public information were suddenly removed. Sources quit talking. Reporters say they have never seen the Pentagon as intimidated about talking to the press as they do now.

When facts are hard to come by, the press, other studies have shown, tends to fill the vacuum with analysis, opinion and speculation.

Even so, when the press was citing sources, a high percentage of them remained on the record, even if they were offering more analysis than strictly facts (76% in September and November, 73% in December).

Some Media Are More Fact Oriented Than Others

Newspapers stuck to the facts more than television. In September, 85% of what appeared in the papers was strictly factual. On television, it was 20 points lower—64%.

Over time, newspapers saw this level of strictly factual accounts decline, but even at its lowest it was higher than in broadcast.

Reporting in Newspapers
Fact 85% 81% 73% 82%
Analysis 10 13 23 13
Punditry 5 5 4 5
Total 100 100 100 100

On television, the mix of fact versus opinion seemed to rise and fall according to the story being covered.

During the collapse of the Taliban in November, for instance, factual reporting fell to less than half of all the coverage (to 46%, down from 64% in September) as journalists and experts speculated about whether the retreat was real, what would happen next, whether al Qaeda would follow.

In December, during the hunt for bin Laden, factualness increased again (to 56%). Journalists and experts seemed more reluctant to guess about the terrorist leaders whereabouts, a matter that could easily be proved right or wrong.

Reporting on Television
Fact 64% 46% 56% 57%
Analysis 18 28 22 22
Punditry 17 26 22 21
Total 99 100 100 100

The Clinton Scandal Comparison

If coverage of the war has become less straightforwardly factual with time, that contrasts to how the coverage evolved during the last great political upheaval the press contended with—the scandal involving Monica Lewinsky and Bill Clinton.

Early Coverage
Clinton Scandal v. War
Factualness 59% 75%
Analysis 23 14
Punditry 18 11
Total 100 100
1 Jan. 21-24, 1998
2 Sept. 13-15, 2001

During Lewinsky, the press initially was condemned for rushing to judgment and then over time became more factual and cautious.

In the war on terrorism, the press was praised for caution at first and has become more interpretive since.

By December, indeed, less of the press coverage of the war on terrorism was strictly factual (63%) than was true six weeks into the coverage of the Clinton-Lewinsky scandal (74%).

The level of punditry in the later war coverage is actually higher than the amount later in the Lewinsky scandal.

The two stories differ in obvious ways. One was sordid and controversial. There were arguments about its importance and appropriateness. The other is an international crisis of undoubted significance.

But the differences in the coverage say something about how the press works. The quick praise of the media's early work may have led some journalists to become less careful. The early criticism in Lewinsky may have caused more restraint later on.

Later Coverage
Clinton Scandal v. War
Factualness 74% 63%
Analysis 18 22
Punditry 8 14
Total 100 100
1 March 5-6, 1998
2 Dec. 10-12, 2001

And the differences in the two events are likely a factor. The Lewinsky scandal occurred in media's home field. The war may have begun there, but it soon moved into mountains and caves in central Asia. Historians have long talked about the difficulty of getting accurate wartime information as "the fog of war."

Nonetheless, one might expect a more factual tenor to coverage of war—a matter of life and death—than that of the ultimate media sport, a political scandal in Washington.

Some of the factual reporting in the Clinton scandal turned out to be inaccurate. While that is also often the case in wartime—the precision bombing and Patriot missiles of the Gulf War proved far less effective than first reported—there is no basis at this point to suggest some misreporting of the war.