When we talk about social networking apps and elections, the question often arises of how online support translates into votes. Now a new study has, for the first time, linked online support on Facebook to actual votes.

Last October, we reported on a study by Bentley College professors Christine Williams and Jeff Gulati that looked at how Congressional candidates used Facebook to reach supporters. At the time -- this was before most presidential candidates had announced, and before the November mid-terms -- the duo's goal was to "determine whether candidate postings on the site are a useful viral marketing tool for generating publicity, campaign contributions, and ultimately votes."

Now, they've completed a much larger study of that race that makes it clear that Facebook and other social sites will play at least a minor role in electoral success in 2008.

Last year, Facebook created a profile for every Congressional candidate, and left it up to them to play with; 32% of candidates for Senate and 13% of House candidates updated those basic profiles.

After the election, Williams and Gulati analyzed the relationship between the votes the candidates received and the support they garnered on Facebook. One conclusion that jumped out at me:

The number of the incumbents’ Facebook supporters and the challengers’ supporters indicates that a candidates’ Facebook activity had a significant effect on the incumbent’s final outcome.

Specifically:

• An incumbent that had twice as many Facebook supporters as another incumbent would have received 1.1% more votes than the other incumbent;
• A 1% increase in Facebook supporters of challengers reduced incumbents' vote percentage by 1.5%.

In simple terms, this indicates that candidates with more Facebook supporters did better at the polls. Just creating and updating a Facebook profile didn't affect things, but gathering supporters did.

The authors make it clear that there is not yet any causal relationship between Facebook profiles and votes; that is, more Facebook supporters doesn't simply equal more votes. Instead, attention on Facebook is probably a product of all-around enthusiasm about a candidate, both online and offline.

"We didn't want to make it look like just getting supporters gets more votes," Gulati told me. "Instead, it says something about how people feel about that candidate. Maybe those candidates had a better logistical organization offline, which was mirrored online. Or maybe people are really enthusiastic about the candidate, which can translate into more positive news coverage."

In other words, if a candidate is popular on Facebook, chances are they're popular elsewhere, which makes it hard to point to Facebook as the vote generator.

"Using Facebook isn't going to make a candidate win or lose," Gulati says, but using the site will help them do better than they would have done without it.

Gulati and Williams will next turn to 2008, which will be the first election in which Facebook has been used by the wider population (in 2006 it was only open to students). And then there's YouTube, MySpace, Eventful, and other sites to consider. We've known that these sites can help us gauge the level of support for the candidates; it will be interesting to see what kind of connection, however tenuous, they will have to actual votes.

UPDATE: You can download a PDF of the entire study here.