Newspapers and blogs: Closer than we think?

David Vaina is a research associate at the Project for Excellence in Journalism.

Back in the mid-1850s, the English philosopher John Stuart Mill argued that a citizenry could not, would not, flourish unless it was nourished by the full spectrum of voices that exist among the people:

It is, however, obvious that law and authority have no business with restraining either [side or sides of the debate], while opinion ought, in every instance, to determine its verdict by the circumstances of the individual case, condemning every one, on whichever side of the argument he places himself, in whose mode of advocacy either want of candour, or malignity, bigotry, or intolerance of feeling manifest themselves; but not inferring these vices from the side which a person takes, though it be the contrary side of the question to our own; and giving merited honour to every one, whatever opinion he may hold, who has calmness to see and honesty to state what his opponents and their opinions really are, exaggerating nothing to their discredit, keeping nothing back which tells, or can be supposed to tell in their favour. This is the real morality of public discussion.

Well over one hundred years later, the blogosphere came into our lives, allowing us, in the words of Jerome Armstrong and Markos Moulitsas Zúniga, to “hear voices that had been shut out of the corporate media outlets.”

These old “corporate media outlets,” refusing to fade away, have held their ground. According to William Dietrich, a writer with the Seattle Times Sunday magazine, the sacred purpose of the newspaper reporter “is to fulfill an essential function of our democracy not just by disseminating information but also by analyzing it, detecting patterns, spotting trends, and increasing societal understanding.” Indeed, bloggers may generate a more democratic Public Square, but can they facilitate a more comprehensive understanding of how political events are most likely to evolve, the Old Guard worries and wonders. In other words, Mill might not be enough.

To contribute to this Great Debate, I decided to conduct a content analysis of how blogs and newspapers covered the Iraq War during one week in late March 2007. By looking at how the two media have sourced their news stories, I hoped to offer insights into what exactly the American public “hears” from newspapers and blogs.

More specifically, my research, by examining five major newspapers and six popular political blogs, sought to answer three questions:

  • Which media platform uses more sources?
  • Which offers a more diverse range of sources?
  • And which types of sources are more prevalent in each platform?


Overall, the data showed that blogs included a higher number of total sources and a slightly wider range of sources.

Blogs included an average number of nine sources per blog posting, compared to an average of just six for newspapers stories.

The gap between newspapers and blogs was considerably narrower when evaluating the types of sourcing. Still, blogs were slightly more diverse in their sourcing, with four sources per posting compared to an average of three in newspaper stories.

Digging deeper, which types of sources were the two media most likely to use?

Both blogs and newspapers were likely to include traditional Washington sources, both political and intellectual.

But blogs and newspapers did diverge in several key ways. Compared to newspapers, blogs were considerably less likely than newspapers to include official Iraqi sources.

And perhaps as a tell-tale sign of what the mainstream press really thinks of the blogosphere, just two percent of newspaper stories used a blog as a source. Not surprisingly, bloggers used other bloggers as sources at almost the same frequency as they used the mainstream press.

Sourcing in Blogs

Seven in ten (69%) blog postings included a mainstream media outlet (e.g. Washington Post, AP, The New York Times) as a source and 64% used other bloggers as sources.

Political Washington was well represented. Thirty percent of all stories had a source from a Democratic politician or party strategist, 28% included one from a Republican or GOP operative, and 23% included a source from the White House.

Meanwhile, a quarter (25%) included sources from the Pentagon, a soldier fighting in Iraq, or an immediate member of a soldier’s family. Ten percent of all blog postings had a source from other government officials, such as analysts from the State Department or the American embassy in Iraq. Furthermore, 16% of all postings included a government document as a source, such as a hyperlink to a PDF of a legislative bill or the complete voting results for a particular bill from the Office of the Clerk at the U.S. House of Representatives.

Considerably fewer blog postings, however, included sources from Iraqi government officials (11%), such as local police and security forces and hospital administrators, and an even smaller number offered sources from Sunni or Shiite politicians (five percent). And only two percent of all postings included a source from the Iraqi insurgency.

Five percent of posts included sources from Iraqi civilians, and eight percent had sources from U.S. civilians.

Finally, a quarter (25%) offered a source from a non-partisan, non-governmental entity, such as a think tank, polling organization, or university.

Sourcing in Newspapers

Turning to newspapers, the most frequent source was a U.S. military official or family member. Over half (53%) of all newspaper stories included a source from this cohort—more than double the percentage in blogs.

The second most common source was a Democratic one; more than three in ten stories (32%) offered a Democratic source.

A quarter (24%) included a source from the Bush Administration, and another 16% had a source from other Republican politicians or strategists.

Another 22% included a source from other government officials outside the halls of Congress, the White House or the Pentagon.

Newspapers were also likely to offer an Iraqi point of view. Thirty-one percent of all stories included sources from the Iraqi authorities. Two in ten (20%) stories included sources from either Shiite or Sunni politicians. An additional seven percent was from sources coded as insurgents.

At the non-political level, newspapers were more likely to quote an Iraqi civilian, with ten percent of all stories offering this point of view. Half that percentage (five percent) included sources from U.S. civilians who were not family members of an American solider fighting in Iraq.

Twenty-three percent used a poll, statement from a non-partisan think tank, or academic as a source.

Finally, eight percent of stories used a mainstream media outlet as a source, and just two percent included blogs.


Much of the current debate in journalism that centers around how sourcing is used in blogs concerns the issues of verification of information not reported in the mainstream press. But for now, this doesn’t appear to be their raison d’etre. The function of blogs may be an equally important one, however, offering a more nuanced, synthesized perspective not found anywhere else on the Web.

Perhaps what’s most at stake for blogs is to evaluate which voices are being synthesized. According to the data for this study, an admittedly limited one, bloggers may be missing perhaps the most important piece of the political puzzle when we acknowledge the realpolitik of Iraq.

Both the American and Iraqi people are growing increasingly weary of the American military presence in Iraq, according to public opinion polls in both countries. If there is one point Democrats and Republicans can agree on it is that Iraq’s future success rests on the further strengthening of Iraq’s political institutions.

Right now, it may be that the traditional press—represented by newspapers here-has picked up on this better than blogs. The data shows that roughly four times as many stories in newspapers included sources from leading Sunni and Shiite politicians as did blogs. Where blogs excelled, with more bloggers, media sources and original texts as sources, is perhaps more easily to duplicate for newspapers on their websites. What cannot be mimicked so easily is the ability to discern which way the political winds are blowing in Baghdad and Washington.

One might dismiss this conclusion as an elitist, Lippmanian one. Regardless, it begs the question of whether or not the public most benefits from a traditional journalist sensibility that, despite its flaws and declining commitment to foreign affairs, can still be found at the country’s best newspapers. Perhaps all those years of having boots on the ground overseas still colors, positively, newspaper coverage.

However, one should keep in mind that only a third (34%) of all bloggers considers their blog a form of journalism, according to a study from the Pew Internet & American Life Project. So my insights may be a case of trying to fit a round peg into a square hole. Furthermore, until the mainstream press can better understand that media consumption and production are increasingly conversational, collaborative activities—where bloggers and citizens talk to each other—perhaps the best advice I can give is to take the time to read a newspaper and a blog or two.

About the Study

For this study, I counted the number of sources over seven days in late March 2007 (March 23-March 29). Only stories with the war in Iraq as the dominant story (50% or more of the story) were coded. Overall, 172 newspaper stories and blog postings–the units of analysis–were coded.

Sources did not have to be original. For example, a blog that quoted an interview from Senator John McCain that originally appeared in the Washington Post would be counted as a source, even though the actual reporting was not done by the blogger. Original sources, though in small numbers, could be found in blogs, most notably in Greg Sargent’s postings on Talking Points Memo.

First, I looked at five major newspapers: Los Angeles Times, The Washington Post, The New York Times, USA Today, and the Atlanta Journal-Constitution. Wire stories that appeared in newspapers were included. A total of 111 newspaper stories were coded.

Second, I conducted an analysis of three major blogs from the left and three from the right. They included: Talking Points Memo, Political Animal (the Washington Monthly blog), Daily Kos, Michelle Malkin, Powerline, and Hugh Hewitt. A total of 61 blog postings were analyzed for the research.

For blogs, a source was defined as those that were available either on the homepage posting or those on secondary pages within one mouse click from the original blog posting. Then, sources within these secondary pages were coded as well (e.g. links to other news sources, bloggers, and government documents). This methodology was employed in order to measure—as much as possible—the total available number of sources that are consumed by the typical blog reader, and not just those that appear in the original blog posting. Sources within tertiary pages (and beyond) were not coded because I felt that only a small number of blog readers would actually read this deep into a blog posting. Nevertheless, these tertiary (and beyond) pages theoretically expand the number of potential sources and should be kept in mind before forming any firm conclusions about the nature of sourcing in blogs.

Online newspapers and the 2006 election: bland ambitions?

Alexis de Tocqueville once characterized American newspapers as a roadmap for citizens, especially as they come together and meet in the public square:

A newspaper then takes up the notion or the feeling which had occurred simultaneously, but singly, to each of them. All are then immediately guided towards this beacon; and these wandering minds, which had long sought each other in darkness, at lengths meet and unite. (Democracy in America, Book Two, Chapter vi.)

As we wrap up another U.S. election, one may ask: are newspapers, in their modern online versions, still meeting de Tocqueville’s great expectations?

Research from the Pew Internet Project shows that the number of Americans who now turn to the Internet for information about campaigns on a typical day has jumped from 11 million in 2002 (the last mid-term election cycle) to 26 million. Surely, a considerable chunk of this 26 million may be visiting national sites such as CNN, NPR, MSNBC, blogs, and increasingly, YouTube. But as the late Tip O’Neill once said, all politics is local, and most online newspapers can offer their readers a uniquely local focus that sites such as CNN and MSNBC cannot.

To better assess the online newspaper industry and its commitment to providing citizens with information on the election, I decided to conduct an inventory of each daily newspaper website in the country from October 14 through November 3, 2006.

According to the 2006 International Year Book printed by Editor & Publisher, there are 1,452 daily newspapers in the United States. Only those online newspaper sites that were free for readers and had fully functioning websites were included, which ultimately reduced the total number to 1,312 sites.


First, I determined if the site had carved out a section of the site and devoted it exclusively to the 2006 election. It had to be a section that was a clearinghouse created specifically for the campaign that a voter would be able to access if he or she wanted news and information just about the election. While virtually all sites integrated news about the election in their general local and national news sections, this alone would not meet the criteria.

Next, I examined the individual components of those sites with special election sections to determine their depth and richness. First, I calculated the number of sites that exhibited multimedia, other than polls. Were there podcasts or video clips of debates, or staff-generated interviews of the different candidates, for instance?

Second, I tallied how many sites offered information specifically on the candidates, and sought to break down the issues for the reader. For example, could visitors read individual profiles of a particular candidate?

Third, I surveyed how many of the sites provided readers with a chance to interact with either reporters or other readers. Were there blogs or forums in which readers could express their own opinions?

And fourth, I determined the number of sites that included details on the logistics of voting: registration information, polling locations, voter-eligibility requirements, and primary results.


I found that the industry’s overall performance can probably best be assessed as uneven. Just 27% of all online newspaper sites offered a separate section for campaign and election news. Digging deeper, a few findings stand out.

First, it may be that the rich are getting richer. Those sites with the largest print circulation, and arguably, the most revenue and resources to allocate for an online election section, were most likely to display one. According to the International Year Book, the average daily circulation of all newspapers in the country is roughly 36,700. Meanwhile, the average daily circulation for those newspapers sites that offered a special election section was more than double that number at 86,500.

Breaking down the numbers by state also shows that those states with the highest concentration of registered voters were more likely than others to have sites with election sections.

Among those states with the highest number of sites that included election sections were Florida (56% of all daily sites displayed an election section), Maine (50%), Maryland (60%), Oregon (59%), North Carolina, (40%), Oregon (59%), South Carolina (40%), Vermont (43%), Virginia (42%), Washington (57%), and Wisconsin (43%). With the exception of Virginia, all of these states had a voter registration percentage equal to or higher than the national average, according to the latest registration data from the U.S. Census Bureau. One might assume that registered voters were already the most engaged citizens while unregistered voters could have most benefited from additional coverage.

On a different note, there was also the possibility that ownership affected the overall number of sites offering an election section. As other research has shown, there is a heavy concentration of ownership among the country’s newspapers. While conducting this study, I noticed that many sites were straight-jacketed by the homogenous online format associated with a particular newspaper owner. Thus, if there was one company that did not offer an election section, all the sites owned by that company did not do so–if they were locked into a general Web structure. The one exception to this trend was Gannett.

The sites fared slightly better when their individual components were measured. Fully sixty percent of all sites with election sections offered in-depth information on the issues and candidates participating in the elections. There were many biographical and professional profiles, and “Q & As” allowed citizens to quickly compare and contrast opposing candidates.

Meanwhile, 56% of all sites with an election section provided readers with the opportunity to interact with stories and other readers. Not only did many sites offer readers the chance to post a comment on a news story or profile, but several sites established campaign-specific blogs for the election.

Next, 54% of the sites with election sections listed information on voting logistics. Perhaps this is the most surprising finding since it involved seemingly so few resources. Many sites simply linked to the local Board of Elections where voters could find their polling place or request an absentee ballot and this was sufficient to meet the criteria for this aspect of the study. Why only a small majority was able to do so baffled me.

And finally, roughly four in 10 included multimedia in their election sections. Since it could very well be that resources were the key factor for online website editors and staff, it is perhaps not surprising that so few sites could offer video, audio, interactive maps, or slideshows for citizens.

Many Americans may be satisfied with the breadth and depth of online coverage of the election. National sites like MSNBC and CNN offer a dizzying array of multimedia and investigative journalism that only a handful of newspaper sites have the personnel and resources to provide. Moreover, there is the 800-lb gorilla–television–that still serves as the overwhelming choice for most Americans when it comes to election news and information. But because newspapers have long been considered the standard bearer of quality and reliable reporting on local politics, particularly as the nature of the web is able to overcome the limits of space that limited print newspapers, it seems rather disappointing that such a small number of sites are able to meet the lofty ambitions set by de Tocqueville over 150 years ago.