Nine concrete, specific things we actually know about how social media shape elections

Nearly a quarter of American time spent on the Internet is spent on Facebook. 15 percent of online Americans tweet. Two-thirds of Americans use a social network.

And there's an election in November.

Opinions about the Internet's potential effects on politics range from the thoughtlessly utopian to the haughtily dour. We don't often get a reality check.

How do social media actually shape elections? What does research say about how social media have affected elections in the past? When you take public discourse and run it through the Twitter spaghetti maker, how does it look? 

Some of these may seem obvious, but the goal is sanity over bloviation. Having looked at the literature, here are the nine things I think we can definitively say about how social media can shape elections, based on the way it already has done so, in elections around the world:

N.B. As always, let's just accept the limits of the phrasesocial media. I'm using it to mean the Web 2.0 Friends: Twitter, Facebook, YouTube, maybe a little Pinterest on the side. All media are social; all media are user-contributed (ditto all "knowledge") and "social media" works well enough to refer to both.

1. People and campaigns mostly use social media for dissemination, not dialogue.

"Neither campaign made much use of the social aspects of social media."

So says the Pew's study, released last week, on how the American Presidential campaigns are using the web in 2012. The abstract of that study describes social media "mainly as a way to push messages out" without having to pass those messages through a journalistic filter. It also noted that the only place where campaigns used non-professional, "citizen" content was in the Obama campaigns's news blog. Everywhere else, professionally-made content reigned. Twitter accounts even rarely or never retweeted ordinary voters:

[J]ust 16% of Obama's tweets over the two-week period studied were retweets. The Romney campaign had just one retweet during this period -- something from Romney's son Josh.

Pew just looked at candidates, though. So to understand how social media shapes an entire election's playing field, you have to look globally.

Many fewer Swedes use Twitter than Americans (someything like one to eight percent of Swedish internet users), but that may make studying the country's tweets easier. In the country's 2010 Twitter elections, Anderse Olof Larsson and Hallvard Moe studied all global tweets tagged with #election2010 (in Swedish, #val2010). They found that dissemination still trumped. A full 60 percent of the tweets were "singletons" -- tweets that weren't an @-reply or retweet. Only seven percent of tweets tagged #val2010 were @-replies. This could be a sign of simply what tweets get hashtagged, but I think it gestures, too, to the fact that most of what happens on Twitter is broadcast.

2. Campaign websites remain the hub of US Presidential campaigns.

This is a Pew finding, and it ties right into how campaigns use social media for broadcasting. Write the study's authors:

Even if someone starts on a campaign's social network page, they often end up back on the main website-to donate money, to join a community, to volunteer or to read anything of length.

They add that this has heightened over the course of the campaign: A July redesign of the Obama 2012 site made it so that rather "than sending users to the campaign's YouTube channel, the video link now embeds the campaign videos directly into the website, where the only videos are the ones Obama wants you to see."

Read more at The Atlantic.