One of my enduring research interests is in the development of blogosphere into an organizing space for communities-of-interest. Blogging is generally treated as citizen journalism, with an assumption that the blog serves as a sort of personal megaphone, echoing out into cyberspace to greater or lesser (usually much lesser) effect. That’s an apt analogy for personal and small group blogs like this one, but the analogy falls apart for community blogs with diary structures, recommendation systems, etc. A site like Dailykos or Mydd has more in common with the Sierra Club than with the Washington Post or a talk radio show. Markos Moulitsas is the vocal leader of Dailykos, but the value of that site comes from the active participation of its tens of thousands of diarists and commenters. There is a difference-in-kind between the “kossack” community and a prominent individual blog like those maintained by Glenn Greenwald and Matthew Yglesias.

Okay, with that throat-clearing out of the way (I’ve got a journal article in the pipeline for those interested more generally in the topic. Shoot me an e-mail and I’ll send you a draft copy), on to the topic of this post: if community blogs function as interest groups, how effective are they? One of the easy metrics in the past has been to follow the money. In the 2005-06 campaign cycle, MyDD, SwingStateProject, and Dailykos supported a joint slate of candidates through ActBlue.com. All told, their “netroots candidates” fundraising page bundled $1.5 million for a total of 17 Congressional candidates. To put those numbers in perspective, 7 of the 17 candidates also received funding from the Democratic Congressional Campaign Committee. Five of the seven got more money from the blogosphere than they did from the DCCC (see below). Let me say that again: the bloggers provided more money than the Party.

2006 Candidate Name BlueMajority Funding DCCC Funding
Paul Hodes (NH-02) $53,900 $28,785
Larry Kissell (NC-08) $62,250 $46,260
Eric Massa (NY-29) $58,950 $12,500
Patrick Murphy (PA-08) $63,882 $64,152
Darcy Burner (WA-08) $79,906 $149,680
Linda Stender (NJ-07) $48,578 $30,478
Joe Sestak Jr (PA-07) $133,573 $72,775

BlueMajority recently broke up – each of the sites wanted to endorse different slates of candidates. It occurred to me that this creates a sort of natural experiment in comparing community blogs. These are very similar blogs with very similar reader bases. In the past, they’ve pooled funds and we haven’t known what percentage came from where. Yesterday marked the close of Q1 campaign fundraising, so OpenLeft (which broke off from MyDD last summer), Dailykos and MyDD were each conducting their own last-minute donor drives. Each of these organizations is among the top 25 progressive blogs (as of November ’07, Dailykos was #1, MyDD was #9, and OpenLeft was #23, check out my article in IPDI’s PoliTech review for methodology). Each gets substantial traffic, hyperlinks, and comments. Dailykos is the mothership, though, the power-law winner, and their participatory activity puts the rest of the blogosphere to shame. So, point is, what sort of relationship is there between various metrics of blog authority and fundraising numbers? Well let’s see…

We’re currently in a lull for political blogs, the eye of the storm between the crazy-long primary season and the general election. Site visits today thus look a lot like they did pre-primary. According to Sitemeter, Dailykos is averaging 746,482 unique visits/day, MyDD is averaging 36,761, and OpenLeft is averaging 15,896. So Dailykos gets about 20.3 times more traffic than MyDD and MyDD gets about 2.3 times more traffic than OpenLeft. Taking a look at hyperlinks, Dailykos has a Technorati authority score of 10,315 (ranked #12 overall), MyDD’s score is 2,197 (#601 overall), and OpenLeft’s score is 1,821 (#948 overall). Usually I like to track comments/week as a third metric of community strength, but I haven’t automated that function yet and it would take 2 or 3 workdays to count up all the comments on those three sites by hand. Anyway, point is that the three are closer by this metric, with dailykos holding a 5-to-1 advantage and MyDD only about 15% more densely-linked than OpenLeft.

Now, let’s take a look at the three ActBlue pages.

Dailykos originally set a goal of 1,000 contributions. They blew past that mark a few times, eventually bringing in 1,734 individual contributions totaling 170,488. That’s an average donation of $98.32.

MyDD originally set a goal of 60 contributions. They edged past that, eventually collecting $3,300 from 62 individuals. That’s an average donation of $52.23.

OpenLeft originally set a goal of of 50 contributions. They did better than expected, with 63 individual donors giving a total of $5,335. That’s an average donation of $84.68.

Now this isn’t a lot of data, certainly not enough to draw any conclusions. But it is enough exploratory data to start raising hypotheses. So here are a few brief thoughts on what these numbers suggest…

-First off, the ratio of dailykos’s contributor totals to MyDD’s and OpenLeft’s totals was about 27-to-1. That’s bigger than the 20-to-1 sitemeter margin (and some argue that Dailykos’s sitemeter stat is almost doubled thanks to a bug in the sitemeter software). Dollar-for-dollar is even more extreme. Comparing the total dollars generated by the Dailykos community to those of the MyDD community, the Kossacks ponied up 51.7 times more dough than Jerome’s crowd (and just under 32 times more than the OpenLefters). My hunch is that the reason for this comes back to a lesson from the power law infrastructure: “It’s Good To Be the King.” Most readers of MyDD and OpenLeft probably also visit Dailykos. When choosing where to donate money, they may lean towards the biggest site. This would be particularly true if they self-identify as part of the “netroots,” because boosting the largest netroots site makes the blogosphere look more effective. If blogosphere-generated money was all broken up into groups of 50 or 60, it would be easier to say that no blog community is all that important. Readers of all three sites are thus nudged toward donations at the biggest site. (full disclosure: I donated my ten bucks through dailykos, even though I prefer OpenLeft as a source of news and discussion. This hypothesis may just be me trying to rationalize my own actions)

-Second, despite lagging in visits/day and hyperlink authority, OpenLeft is now raising more money for candidates than MyDD. This is a bit of a surprise. Granted, the numbers are small enough that we shouldn’t jump to conclusions too soon, but it suggests that we might want to toy with a research design that includes deeper content analysis of the subject matter on the blogs themselves. MyDD, for instance, has spent months as a hotbed of Clinton supporters, while OpenLeft has provided more dispassionate analysis through its “Race to the Nomination” series and its analysis of competitive Senate seats. Anecdotally, it seems like they’ve been picking up defectors from the OpenLeft community who got tired of the bickering over Clinton and Obama and wanted heightened discourse. The higher level of contributions could indicate that OpenLeft is attracting more serious political adherents (or it could just be noise, or luck).

-Third, lets put the raw numbers themselves in perspective. The Dailykos community has just as an identifiable community put 170K behind their favored candidates, and that’s just through June. That’s far more than the Sierra Club or AFL-CIO can directly contribute to a set of candidates, thanks to the difference between bundling money and PAC donations. It’s already more than most organizations will put into congressional races this cycle. Not to demean the contributions of MyDD or OpenLeft, that’s still serious money they’re raising, but it’s on a manageable scale. $5,000 isn’t enough to elect someone to Congress, particularly since it’s divided between 7 candidates. $170,000 is party-sized contributions, though. It sets Dailykos apart from the rest of the blogosphere, and head-and-shoulders above offline political associations.

That’s all these numbers are saying to me. Like I said, they’re very early numbers, hypothesis-generators more than theory-testers. Anybody have any thoughts on what they mean?