Blog paper

by Henry Farrell on July 21, 2004

The paper that Dan Drezner and I have been writing on political blogging is now fit, more or less, for human consumption – it’s available “here”: We’re going to present it at APSA where we’re organizing a “panel on blogging”: We’re grateful for comments, suggestions and criticisms – this is only a first draft.

The key arguments of our paper:

(1) Blogging is politically important in large part because it affects mainstream media, and helps set the terms of political debate (in political science jargon, it creates ‘focal points’ and ‘frames’). Note that we don’t provide an exhaustive account of blogs and politics – some aspects of blogging (fundraising for parties, effects on political values in the general public), we don’t have more than anecdotal data on. There’s plenty of room for other people to do interesting research on all of this.

(2) Incoming links in the political blogosphere are systematically skewed, but not according to a “power law” distribution, as “Clay Shirky”: and others have argued of the blogosphere as a whole. Instead, they follow a lognormal distribution.[1] We reckon that the most likely explanation for this is that offered by “Pennock et al.”: – they argue that not only do the ‘rich get richer’ (i.e. sites that already have a lot of links tend to get more), but that link-poor sites stand a chance of becoming rich too. Late entrants into the political blogosphere can do well as long as they’re interesting and attract some attention – bad timing isn’t destiny.

(3) Because of the systematic skewedness of the political blogosphere, a few “focal point” sites can provide a rough index of what is going on in the blogosphere – interesting points of view on other sites will often percolate up to them as smaller blogs try to get big blogs to link to them, by informing them of interesting stories. Thus, we may expect that journalists and other media types who read blogs will tend to all gravitate towards a few ‘big name’ bloggers as their way of keeping up with what is going on in the blogosphere as a whole.

fn1. For which we’re grateful to “Cosma Shalizi”: – when we realized that we weren’t dealing with a power law distribution (the log-log relationship looked dodgily curvilinear), he not only suggested alternative distributions and how to test fit, but actually volunteered to do the tests himself.



praktike 07.22.04 at 12:18 am

Wonkette’s on the panel?

That ought to be fun.


Detached Observer 07.22.04 at 12:24 am

“Late entrants into the political blogosphere can do well as long as they’re interesting and attract some attention – bad timing isn’t destiny.”

You mean to say the reason people dont read my blog is because its boring?



Brian Weatherson 07.22.04 at 12:34 am

Just from untutored observation it seems to me that while bad timing might not be destiny, good timing can be a large part of it. The first movers in any field seem to stay near the top, even when there’s little evidence they are in any way better than the competition. (Exhibit A here is my philosophy blog, which still gets way more hits than some of the more recently formed philosophy blogs which have been, at least over their lifespans, much more interesting.) Is there any way to confirm or refute this speculation from the data?


glory 07.22.04 at 2:10 am

how about the tendency toward polarisation among political bloggers? :D



John Quiggin 07.22.04 at 2:30 am

A point you haven’t really stressed in relation to the question “why do journalists pay attention to blogs” is that a major component of blogging is media criticism, including both general debates about media bias and critical comments on particular stories and articles.

Apart from the occasional letter to the editor, the average journalist doesn’t get a lot of feedback in the traditional course of things, but journalists, and particularly pundits, get a heap of attention from blogs, both critical and favourable. It’s hard to imagine that they would ignore this.

The Raines episode is not an ideal example here, I think.


Insight 07.22.04 at 2:55 am

That post is the examplar of how somebody should post for academic research on the blogosphere. Short, but no pandering – this is a market idea that you guys have struck gold on, and you’ll do quite well.


Mike 07.22.04 at 4:20 am

When there is no interaction (or just random interaction) the result is log-normal distribution. As people pay attention to one another so that the probability of “getting richer” starts to depend on current “wealth” level, the distribution becomes more fat-tailed (power law). This is why we see log-normal for the (new) subcategories and power-law for the whole system.

Thus, it seems that political blogs are a “new country”, where a lot of exploration is still possible. In few years, when people will “know” who is who, I expect to see a power law distribution. Can you get access to historical data about political blogs to study their evolution?

“so that our primary hypothesis – that the political blogosphere has a highly skewed distribution of links – is supported.”

Assuming rich do not get richer, we are dealing with a multiplicative process Y=X1*X2*X3*… that will converge to log-normal (central limit theorem), so it will be skewed from the start. The questions you might want to answer are whether there is convergence towards power-law and what does it imply.


David Brake 07.22.04 at 10:12 am

You say, “survey research suggests that the demographics of bloggers do not differ in any appreciable way from Internet users as a whole. One online survey suggests that in terms of gender balance and income distribution, the community of bloggers is more representative of the general population
than Internet users.”

Please provide a citation for these surveys – the results are rather surprising to me.

You state, “Many bloggers desire a
wide readership” – in one sense this seems obvious but do you have any evidence?

You need some explanation of how the “blogosphere ecosystem” database is formed.

Saying “Pennock et al. predicts a roughly lognormal distribution” may be a little over-simplistic – if I understand correctly it allows for some power law distributions in some cases like company homepages.

You say, “they will also often have an incentive to contact one of the large ‘focal point’
blogs, to publicize their post. The latter may post on the issue with a hyperlink back to
the original blog, if the story or point of view is interesting enough” – but this makes the process seem almost automatic and doesn’t take into account the power of these individuals to make choices based on personal criteria (which is a point made later in the paper by Volokh and Reynolds).

For all that I would like to say I thought it was a very interesting paper and I thought the point you made about the way in which journalist bloggers made it easier for other journalists to take the blogosphere seriously is a crucial one and your survey of journalists is very valuable. Since that is an important new piece of empirical evidence can I suggest you put it in the abstract?

It might help explain why the political blogosphere is almost absent here in the UK for example.


Amardeep Singh 07.22.04 at 4:14 pm

Thanks for the post and the article (though I have to admit the statistical math is beyond me).

My question: what about bloggers who have significant numbers of readers who are not themselves bloggers? These people do not have significant numbers of incoming permanent links — word of mouth about their blogs travels generally via email and chat.

I’m thinking in particular of the political blogger Jivha, who was until recently doing a (hyper) active blog in Bangalore, India. He was getting thousands of hits a day, and had (has) 350 inbound links according to Technorati (pretty respectable). But judging from the comment boards, the overwhelming majority of his readers and participants did not themselves have blogs.

Is it possible that differential conditions might lead to different distributions of blogger capital?
In some cases, the issue might be access to the internet. But also, a blogger who takes on a unique subject might draw readers from outside the blogosphere.


Brey 07.22.04 at 4:19 pm

I’m interested because I don’t thikn most bloggers have any effect on the mainstream media. It really seems like the only ones that do are written by people within the mainstream media or regulars on talk shows (i.e. Josh Marshall, Juan Cole). I don’t think that even Crooked Timber has made an impact.


eudoxis 07.22.04 at 10:06 pm

The paper provides a great snapshot of the blog network (blognet? – anyone else irritated with the awkward and plodding “blogosphere”?) , and gives some insight into the connection between it and traditional media. I wonder if the connection between other media and blogs really represent frames. I’m inclined to think that some nodes represent shining beacons, both for readers, other blogs, and traditional journalists. The world of blogs is quite small and a few voices carry an enormous influence that is not represented by the number of network connections, but by the influence such a connection have on other nodes (strength of connections). The echo-chamber tends to involve a few other strong bloggers, but not the readers. From that perspective, the Lott example may not have much support. My impression was that Lott was expendable in the way that Daschle is; he just wasn’t very effective (i.e., an inside job with a convenient origin). The nature of new blogs and their entry into the (cough) blogosphere, is very interesting.


graduate bum 07.23.04 at 12:16 am

how come the paper titles aren’t listed by the panelists names? or am i just not seeing straight?


Pat in NC 07.25.04 at 7:10 pm

I started out reading the in order to find some information on Iraq. I followed links and more links. I have corresponded with others via their email and read their blogs. This has provided a far different view from that of the main stream media. So I started my own blog.


Ron Wright 07.25.04 at 11:42 pm

The media as we know it may very well be replaced by bloggers for the reasons I discuss in two recent pieces:
(2nd reply under this post)

Ron Wright, Moderator
HSPIG Forums Site


Ron Wright 07.25.04 at 11:58 pm

The media as we know it may be replaced as the main source of information by the Blogosphere.

See two recent pieces on our site:
(2nd reply under this topic)

The blogosphere just may be the key element necessary to implode repressive regimes. The free flow of information now transcends political boundaries and without the constraints and biases of media editorial and conglomerate boardrooms.

The endless power of the truth and objective information is what is needed to crush the failed ideology of Islamofascism we are at war with. The power of the great lie no longer controls. Final solutions can no longer occur in secret.

Ron Wright, Moderator
HSPIG Forums Site

Comments on this entry are closed.