Saturday, August 30, 2008

Are Political Blogs Predictive?

Published by Kevin J. Dooley in Managing Technology

2008 will be remembered for the classic battles between Obama and Clinton and McCain and Obama, but political wonks will also note the historical nature of this presidential campaign because of the profound impact that the Internet and social media have had on the dynamics of the race.

From Barack Obama's grassroots fundraising to Ron Paul's "Revolution," bloggers have advocated and vetted candidates in a public manner not seen in any previous campaigns. Involvement in social media allows voters to influence and be influenced.

A reasonable question is: Are political blogs predictive? Specifically, do blogs process and spread information more rapidly than mainstream media (MSM)? Is the blogosphere a "first mover" or "early adopter"? Do blogs create buzz around something that would otherwise go unreported by MSM? In other words, are they a "buzz creator"?

Blogs and nascent stories

From the perspective of the consumer world, we have numerous examples where the blogosphere, or the Internet in general, kept a story alive until MSM could get around to reporting it.

Perhaps the most famous example involved the Intel Pentium chip flaw. When small computational errors were found in the chip and made public in discussion board forums, Intel (and MSM) ignored it. But because of the compounding concern about the flaw, other industry discussion boards and magazines began to pick up the story, until eventually the story broke in the mainstream. It is interesting to note that many execs at Intel, including supposedly Andy Grove himself, were not aware of the issue until it hit MSM.

In these cases, blogs have the potential to give a worthy story legs and sustenance until it "matures," and is ready for diffusion in MSM.

The second situation is one where bloggers take a more active role in actually uncovering news and making it public. The Dan Rather-Bush draft story is an example where bloggers quickly vetted and rejected Rather's story as untrue, and the backlash from the embarrassing event led to the "death" of the Dan Rather brand.

Another example is the August 2006 banning of Coca-Cola and Pepsi by two Indian states after the Center for Science and Environment in New Delhi said it found pesticide residues in product samples that were 24 times above the limits set by the Bureau of Indian Standards.

Most people in the U.S. heard about the ban on August 10 by reading their local newspapers. The story had hit large U.S. MSMs (USA Today, BBC, ABC) two days earlier. The New York Times led the second-wave of MSM with an August 7 report. A first wave of MSM coverage had passed by ignored around August 3, with reports from BBC, Reuters, and the Toronto Sun.

But the Indian blogosphere had been discussing this as early as mid-June. Blogs such as World Prout Assembly and Mission and Justice were carrying information a month before the court ruled against Coke and Pepsi. Finally, if you had been tracking Indian blogs all year long, you would have known that Indian citizens in certain towns were furious with Coke for "stealing" the water from their aquifers, and that trouble was brewing as early as March 2006.

So in the consumer world at least, bloggers can break news, keep news alive until the bulk of MSM gets around to paying attention, and help news spread faster than it otherwise would. How have political bloggers acted as disseminators and creators of news? In my experiences of managing and blogging about Wonkosphere, a web site which tracks the candidate's buzz using patented natural language processing technology, I have noticed the following about the political blogs we track.

Blogs and the political process

First, political blogs are important to the political process because they surface more news and opinion, and disseminate it more rapidly, than MSM. This can be both a threat and an opportunity. Second, campaigns use the political blogosphere as a laboratory to test the stickiness of the candidate's messages, about themselves and the opponents. Finally, MSM looks to the political blogosphere as an "early adopter," and thus the political blogs, by paying attention to certain things and ignoring others, drive a part of MSM's agenda.

In order to understand the role of bloggers as early adopters, I turn to the work of Everett Rogers and his classic book "Diffusion of Innovations", which is the bible concerning how both tangible innovations (like a product) and intangible innovations (like an idea) diffuse in a society.

He identified that there are different "types" of people per when they adopt: first the innovators adopt the new entity, then early adopters, followed by the early majority, late majority, and finally laggards. In Rogers' theory, early adopters are the most important group in terms of the likelihood of broad diffusion, as they take the bold ideas from the innovators and mold them for adoption by the pragmatic early majority.

In a political context, Roger's model implies that political ideology is first adopted, or not, by political bloggers who act as both innovators and early adopters. In turn the rest of the voting public reads and is influenced by bloggers’ opinions, thus bloggers become a secondary communication medium for the candidates, either supplementing or competing against the candidate’s own communications.

Will bloggers vote in the same way that the general populace does? According to Roger's theory, the answer would be yes. As people at the front-end of the adoption curve, bloggers have to "adopt" before the general populace does.

Not all bloggers adopt the same idea -- as ideological leaders, their views are going to tend to be more intense and diverse, and more ideological and less pragmatic. Nevertheless, bloggers in the aggregate will tend to move ahead of the curve and thus be predictive. Indeed, statistical analysis of the Wonkosphere Primary Campaign data indicated that changes in buzz share in both the conservative and liberal blogospheres were predictive of changes in voter polls two weeks later.

Imagine a funnel of political ideas. At the front-end of the funnel, many ideas exist in an ideological soup. The campaigns first use bloggers as innovators, listening to their feedback to hone their messages and frames, and then other bloggers in turn act as early adopters and help select and shape those ideological innovations in such a way that they are attractive to the early majority, i.e. the primary voter.

Supplements rather than substitutes

Finally, how are political blogs the same or different from MSM? Through Wonkosphere, we have noticed that political blogs are consumed in much the same manner as mainstream media is, which indicates that readers treat political blogs not as separate from, but rather as part of, mainstream media. Wonkosphere traffic is greatest on Monday, and tends to peak before breakfast, lunch and dinner, i.e. when people are cruising on the net to end a portion of their work day. Blogs act as newspapers for most.

Second, very few blogs break stories. From our data, the vast majority of bloggers still rely on mainstream media for the content they comment on. In fact, a blogger is just as likely to cite mainstream media as they are another blogger. Thus, bloggers are primarily amplifiers rather than sources of news.

Third, the popularity of political blogs tends to follow a Pareto (power) law, meaning that there are a few blogs that have hundreds of thousands of readers while most blogs only have a handful of readers. This means that the influence of blogs is distributed in the same way, leading to the development of elite blogs (MyDD, Townhall), in the same way we have elite mainstream media sources (New York Times, Newsweek).

Put together, these patterns imply that political blogs are acting as supplements to mainstream media, rather than substitutes for it. Their impact on the system is to increase volatility: blogs make most news spread faster, but sometimes they slow it down. Blogs spread fact, opinion, truth, and slander more rapidly -- it is not biased in that regard. Only a few blogs influence opinion most of the time, but any single blog has the potential to impact everyone. That’s why this year’s race is particularly exciting—will something break in the blogs that ends up being the tipping point?

-- Kevin Dooley, a professor of supply chain management in the W. P. Carey School of Business, is a world-known expert in the application of complexity science to organizations. He has co-authored two patents concerning Centering Resonance Analysis, a novel form of network text analysis, and is Chief Executive Officer of Crawdad Technologies, LLC. Crawdad provides text mining software and services for applications in marketing and academic research.

No comments: