A 2005 study by the International Conference on Knowledge Discovery and Data Mining described the state of the political blogosphere in 2004 Presidential elections, at a time when this medium was beginning to be important. It was a period when 9% of Internet users categorized themselves as ‘frequent’ or ‘sometime’ readers of blogsites, a number significant enough to warrant attention. Howard Dean famously issued his information bulletins through them and “the Democratic and Republican parties further signaled the established position of blogs in political discourse by credentialing a number of bloggers to cover their nominating conventions as journalists.” However tentatively, the political blogosphere had arrived. And it was bifurcated, it seems, at birth.
The authors of the study divided their samples into the “conservative” and “liberal” parts of the blogosphere, depending on the positions they took on certain litmus-test issues, like gun-control, abortion and the like. Having found a way to assign colors to each dot, the researchers then examined how these blogs linked to other sites. They found curious differences in the way the conservative and liberal blogosphere were topologically organized. They resembled each other in that the respective “camps” tended to draw their links from like minded sites, but there were differences too. The liberal blogosphere was somewhat more insular than the conservative blog universe.
a great majority of the links remaining internal to either liberal or conservative communities. Even more interestingly, we found differences in the behavior of the two communities, with conservative blogs linking to a greater number of blogs and with greater frequency. … In our study we witnessed a divided blogosphere: liberals and conservatives linking primarily within their separate communities, with far fewer cross-links exchanged between them. This division extended into their discussions, with liberal and conservative blogs focusing on different news, topics, and political figures. An interesting pattern that emerged was that conservative bloggers were more likely to link to other blogs: primarily other conservative blogs, but also some liberal ones. But while the conservative blogosphere was more densely linked, we did not detect a greater uniformity in the news and topics discussed by conservatives.
One of the interesting questions is why this division persists in the face of a clear commercial incentive to be “moderate”. Common sense suggests that sites which bridge disconnected parts of the Internet (at the collision point of these two memetic clouds) may get more traffic than those which remain in their own little corner. It is almost always better to be at the “crossroads” than in the cul-de-sacs. A glance at the graphic above shows that if you are far enough left or far enough right then nobody much will link to you. And indeed, the bigger purple or red grapes (presumably representing the bigger sites) seem visually concentrated near the center. It’s been said that newspapers in the past attempted to adopt a magisterial tone in order to attract a mixed readership of liberals and conservatives, thereby broadening their audience, and consequently their ad revenues. Why have the blogs divided into two camps instead of merging into a single blob? One possible explanation is that many smaller bloggers do not operate their sites for profit and therefore political preference and ideology are far more powerful drivers than any small amount of revenue that added traffic might bring. In other words, since most bloggers don’t blog for a living, they will write what they please instead of trying to optimize their web ad revenue.
It would be interesting to see how much things have changed in four years. Jason Lee Miller of Webpronews argues that despite assertions to the contrary, the “conservative blogosphere” is bigger than its liberal counterpart, though I suspect that depends on how you measure things. However, there is apparently some real likelihood that the role of the political blogosphere is bigger than the 9% number given in the 2005 study.
To say the Internet skews left, as many believe, might be a little deceptive. True, The Huffington Post is the most popular of standalone sites, grabbing 4.5 million visitors in September, up a whopping 472 percent from last year.
But combining the second and third place in the top three, the audience is nearly even with HuffPo with over 4.4 million visitors split between Politico.com, up 344 percent from last year and attracting nearly 2.4 million visitors, and The Drudge Report, which was up 70 percent to almost 2.1 million visitors in September.
If you combine the top 15, it’s conservative blogs bringing in the largest overall audience. Left-leaning sites grabbed an audience of about 6.6 million in September, while right leaning sites attracted a combined audience of 8.4 million. It may be because political blog readers tend to be more affluent.
I’m not so sure affluence has anything to do with being conservative or liberal. And those numbers, it seems to me, are way too low and I think they are wildly inaccurate. But Miller, in mentioning affluence, touches on something that has not yet been closely studied: the effect the progressive enlargement of the blogosphere will have on the left-right divide itself. I think the early days of the blogosphere were perceived to be dominated by the conservatives because they were early adopters. They were early adopters because, unlike most of the liberals active in the culture wars, conservatives did not have platforms in the traditional mediums: print, TV, video, etc, with the exception of radio. So they turned to the blogs. But as the blogosphere matured and the traditional mediums declined, liberal media people migrated in ever large numbers to the Internet. What followed was a transfer of resources from traditional mediums to the blogosphere. Newspaper columnists got blogs; people in broadcast made videos or start video blogs, etc. Eventually the liberal transfereees began to catch up with the conservative early adopters. Moreover, many of them were professionally trained in the media skills and this added to the perception of liberal predominance.
In fact maybe the liberals did predominate for a brief window of time. But maybe that time is now already past. As even newer entrants who were never in the “culture wars” before come online via Twitter, Facebook — and the blogs — the topology is bound to change. The blogospheric significance of events in Iran and China is that people we don’t even know about are going to make an impact on news and opinion in ways that Dan Rather could not have conceived. I think the phenomenon of Internet publishing is migrating away from the traditional culture warriors downward and outward. Downward to nonculture warriors (like the Tea Party amateurs) and outward to international locales. The reporter and pundit of tomorrow is someone who doesn’t even know he is a reporter or is unaware he is a pundit. It’s interesting to consider what this means to people who aim to politically mobilize the average Joe in the near future. The major drivers of the democratization of the Internet have not been content providing sites like the Huffington Post, nor extensions of traditional PR activities like “accrediting” bloggers, but architecture; architecture which enables content provision. In this year of the Iranian demonstrations the Nobel Peace prize should be awarded to Twitter, Facebook and Blogger. Time magazine should consider them candidates for the Virtual Men of the Year, and put Time Magazine itself on its obituary pages. Perhaps the long term political equilibrium of the blogosphere will converge to the underlying distribution of political allegiance in society as a whole. It will certainly be interesting to see what kind of map of the blogosphere researchers will draw in 2010.