This is a prelude to a longer post on the internet in campaigns. I was stimulated by the following passage in Garrett Graff’s Washingtonian magazine piece, which is derived from this book (reviewed here in the New York Times):
bq. There’s an argument to be made that YouTube, the online video site now owned by Google, delivered the 2006 US Senate elections to the Democrats; two GOP incumbents, George Allen and Conrad Burns, went down to defeat after they came under the attack of embarrassing YouTube videos. While Virginia senator George Allen faced a particularly devastating situation with his “macaca” comment, none of the 2008 candidates will escape the wrath of YouTube.
This interpretation of the Virginia Senate race is nothing new, of course. But such interpretations are key to buttressing the notion of a “Youtube election,” which suggests that Youtube clips are consequential to election outcomes. But are they?
Here are the Virginia Senate polls, courtesy of Pollster.com:
Allen used the word “macaca” twice on August 11, with the media coverage beginning soon thereafter. From the graph, it is clear that Allen’s lead began to dwindle at this point. By the beginning of September, Allen and Webb were in a statistical tie.
But then, things changed. Allen came up with fresh attacks on Webb — e.g., on his view of women — and the gap widened again (perhaps because of these attacks, perhaps not). The gap closed to a statistical tie only at the very end, weeks after macaca.
Thus, the “Youtube cost Allen the election” story needs to be complicated.
1) If we assume that it was “macaca” that caused Allen’s lead to dwindle, why should we credit this incident for his loss and not the other campaign stories and tactics that occurred in September and October? Why aren’t we asking why Webb closed so strongly? Is the temporary shift in the polls (perhaps) occasioned by “macaca” really responsible for Allen’s defeat? It’s not even clear, and we haven’t even talked about Youtube yet.
2) What about the role of more fundamental factors that influence election outcomes? Could Webb have won if President Bush’s approval rating had not been languishing at 35%? And what about Webb himself, his qualities, qualifications, etc.?
3) Was it the Youtube clip that caused the effect, or was it the mainstream media coverage of the incident itself? While the two versions of the clip on Youtube have been watched 650,000 times, that figure obviously doesn’t tell us how many unique Virginia voters saw the clip between August and November. Moreover, the number of people who consume traditional media sources — whether in print, on TV, or on the internet — is vastly larger than the number who watched the Youtbe clip. It’s much more likely that people were exposed to the macaca incident through such dinosaurs as the newspaper and the evening news than through a first-hand viewing of the clip itself.
A rejoinder: Yeah, but wouldn’t Allen have won if not for macaca? The answer: we don’t know. And we can’t know. My question is why we should privilege explanations centered on macaca or Youtube more than explanations centered on “fundamentals,” on Webb, or on later campaign dynamics.
Allen’s defeat was not solely, or even primarily, a Youtube phenomenon. The shorthand history, which elevates Youtube above all other things, is wrong.
More generally, it is important to qualify the claims being made about the Internet’s power to affect elections, a question that I will explore in a near-future post.