Bias Disclosure: I'm a yellow-dog democrat, solid ABB, and
presently backing Dean.
There's a lot of speculation in this campaign about grassroots
and its effect. Indeed, this election cycle may be viewed as a
critical turning point in the power dynamics of American politics,
due primarily to uptake of new technology by the various campaigns.
It would be nice, however, to be able to experimentally measure a
candidate's grassroots, particularly with regards to candidates such
as Clark where there has been much debate as to the nature and depth
of his grassroots.
Some questions that have been asked, which would be useful to answer:
How large is candidate X's grassroots? Will candidate X's grassroots
actually vote? Do candidate X's grassroots have depth? Ethnicity?
Regional breadth? Are they insiders or outsiders? How effective is
their community? How much does the campaign listen?
I propose that analysis of participation in an official blog is a
good indicator of grassroots participation, and that answers to these
questions based on blog traffic may reflect ground truth accurately.
My hypothesis is that these blogs are serving as the key new
mass-listening technology which is enabling the new grassroots seen in
this campaign.
As a test, I summed the number of blog comments for all 10
presidential candidates on 10/28/03. In addition, I chose an
open thread on each site and read the tone of the commentary to
try to get a sense of the depth and vigor of the grassroots.
My results were as follows:
Blog Participation for October 28th:
- Braun: No Blog
- Bush: 0 (No comments allowed on "Blog")
- Clark: 1442
- Dean: 1626
- Edwards: 170
- Gephardt: No Blog
- Kerry: 393
- Kucinich: No Blog
- Lieberman: No Blog
- Sharpton: No Blog
Personal Analysis of Community and Tone (non-scientific)
- Clark: Lots of participation, though there was more interpersonal chat
than other blogs. Mostly positive discussion.
- Dean: No surprises here. Big and vibrant with a wide diversity of posting, mostly constructive discussion.
- Edwards: Small, but alive. Lots of pissiness at the lack of message uptake, but constructive discussion on remedies.
- Kerry: Even a casual scan reveals that a few prolific posters are responsible for most of the posts. Lots of anti-Dean and troll-wars.
There were two big surprises for me in this analysis. The first
was the high amount of posting and vivacity of discussion on Clark's
blog, particularly given the "Clark is fake grassroots" memes floating
around out there. This calls those memes into serious question, although
the prevalence of chatter prevents their complete dismissal.
The second big surprise was that of the field of ten candidates,
there were only four blogs (I count Bush as a newsfeed rather than a
blog, given the prohibition of comments). I was particularly surprised
by the lack of a blog for Kucinich and Gephardt, given the
Progressive/Activist base of Kucinich's support and Gephardt's overall
strength. Braun and Sharpton, on the other hand have so little money
that their lack of blogs is no surprise. Lieberman and Bush I can
only fall back to conventional wisdom of "not getting it", which is
unsatisfying at best.
Some projects of immediate interest, based on these preliminary
results: better analysis of the degree of engagement between campaign
leadership and blog community. This is a particularly important
question with regards to Clark's blog. Additionally, it would be
interesting to know if there are "pseudo-official" blogs for any of
the other candidates, where the grassroots gathers without official
sponsorship of the campaign. I am particularly interested in this with
respect to Braun, Sharpton and Kucinich, the candidates which have
strong marginal followings.
On another note, if we are truly interested in transforming the
current political discussion, then it would be a very good thing for
volunteers to offer a working blog to the campaigns which are without
one. Whether or not the blog organizers support a particular
candidate, grassroots organization on a blog appears to have been
highly beneficial in engaging more people in the political process and
in engaging the candidate with the people who support him or her. I'm
not a blog expert, otherwise I'd be offering my services to Gephardt,
Kucinich, Braun, Sharpton and yes, even Lieberman.
I think that an ongoing project of automated analysis to chart the
posting activity for the blogs would be a valuable piece of
data to add to the discussion of grassroots politics. I'm interested
in constructing such a tool, and would be interested in collaborating
with someone who has experience with web-scraping. My own experience
is primarily in AI systems, which can provide the automated analysis
backend.
Some further targets for automated analysis:
- Dialog Categorization (e.g. personal chatter vs. reports vs. policy critique etc.)
- Troll Density and Response
- Regional/Ethnic/Party Diversity
- Astroturf Detection
- Conversational Tone (e.g. defensive vs. constructive vs. reactive)