I offered these remarks regarding my post from last Novermber on traffic differences among left- and right-wing blogs:

Prior to last fall's launch of Pajamas Media, most of the big political blogs ran Blog Ads. Going to the the Blogads site it was surprisingly easy to find and download weekly page views and other data for those blogs. So easy, in fact, that I undertook a quantitative empirical study of the effect of page views, the number of ads, and political orientation on the price blogs could charge.

I posted my summary of that analysis on my blog the same week that Pajamas "went public". The post was picked up by BusinessWeek Online and a handful of bigger blogs. Among the more interesting findings was this: while there was not a signficant difference in the number of weekly page views at conservative and liberal blogs, the distribution of the page views was: Conservatives and liberals predominated in different regions of the spectrum.

The impact that this uneven distribution had on ad price and revenues was notable and pronounced: even when controlling for other traffic, liberal blogs were able to earn more, on average, than their right-of-center counterparts.
A follow up study to determine the effects in the post-Pajamas era is already underway. Drop me a line if you want to know more.

Wretchard offered this explanation for the results of my analysis:

Great post and analysis. I think part of the reason for what you observe is a difference in the structure of the Left-Right blogosphere. Although it would be oversimplifying thing so to say it, the ROC blogs are individual voices -- what someone called with some hint of disparagement the "blogger kings" -- while some of the bigger LOC blogs are actually communities. Like the Daily Kos.

The higher daily page views are only attainable, I think with group blogs, because blogging is a game where the users generate a lot of the content. Comments, groups blogs, etc will bring in their niches to view threads and subthreads. In blogging more than anywhere else, the Long Tail prevails. One other consequence is the differential influence LOC and ROC blogs have. The LOC blogs, because they are in effect, minipolitical parties, immediately translate into party influence. Exhibit A: the Daily Kos.

On the other hand, the biggest ROC blogs really function differently. They are meme generators. Only a few, if any, are linked at the hip to the Republican Party. There have been calls for the ROC blogosphere to become more Kos-like. But I think the reason this is hard is because their constituencies are also fundamentally different. In consequence the blog substructure will be correspondingly different.

I think this going to change because events will increasingly make it both possible and necessary to link up analysis to action. You are already seeing blogs creating their own network of correspondents; serving as enablers of demonstrations, free-whoever type campaigns, etc. And this can only increase into the future. More I can't say except that you folks, not I, am the Belmont Club, in a fundamentally literal sense.

Commentary

Thanks for the thoughtful reply. I agree with your conclusions about the differences in the structure of the left and right-hand sides of the blogosphere. Though I did not explicitly takes this into account, there does seem to be a greater number of lone gunment, solo acts, and one-man bands among the ROC blogs. The LOC blogs surely has its share- Steve Gilliard and Oliver Willis come immediately to mind- but there do seem to be more group blogs among this part of the sample. The opposite appears to be the case for the top ROC blogs, e.g. Hewitt, Malkin, Instapundit, etc.

Another important and related factor to consider is the network structure of the two camps. As I wrote in "TTLB, Open TrackBacks, and Social Network Analysis"

...there is a tremendous amount of research and theory in the area of social network analysis that cries out for application to the blogosphere. This is especially so for that research that examines the network structure of links, i.e. not the number so much as the (relative) position and function of certain nodes within networks. Let me provide a practical example to make this last point clear.

Imagine a simple "ring" network comprised of six blogs where one of them is linked only to its immediate neighbors, as below:

Now imagine another where every blog is linked not only to its neighbors, but to every other blog in the network. This is what is often called a "clique":

Finally, picture a situation where one blog is in the center and that it is linked to all others but those others are linked only the center. I'll call this the "Hub-and-Spoke."

I have a hunch that the right and left-of-center blogs not only have a different internal structure (single vs. multi-authored), they also have different network structures. In other words, they are linked together, as a community, in an altogether different manner. I am not yet in the position to say exactly how they differ, but it's hard for me to imagine that they do not.

SNA has proved an enormously effective tool in analyzing terrorist networks and, as Wretchard noted in 2004, catching Saddam Hussein. I have little doubt that it will prove a useful method for understanding the relative performance of blogs, as well. The major challenge I face right now is finding the right software tools to determine which blogs are linked to one another. If any reader knows of such programs, please advise.