When the Obama campaign created a Facebook-based app in 2012 that would scour users' friend lists to figure out who most needed to be contacted prior to Election Day (your friends in Florida and Ohio, for example), privacy groups were not excited about it. The app compared friend lists from Facebook to publicly available voter lists, which, Electronic Privacy Information Center associate Director Lillie Coney told the National Journal, "widely crosses the line" of protecting privacy. (That Florida friend could be a non-voter because of a prior felony, for example, which he'd rather you not know about.)
News broke Monday that Facebook would turn off the key function that made the Obama app work. (It had been announced by Facebook in the Spring.) No longer will campaigns be able to skim your friend lists as that app did, a development that will make relationship mapping by campaigns slightly less easy than it was two years ago — but not impossible. Facebook has proven the utility of online social networks in increasing turnout. It just prefers campaigns to use its data in ways that help its bottom line.
Facebook is by far the most popular social network in the United States. As of last September, Pew Research figured that 71 percent of online adults use the service, 50 percentage points higher than the second-place service, LinkedIn. "Facebook users are much more politically engaged than most people," Pew reported. And that politically engaged base overlaps more neatly with the Democratic base than the Republican one. Analysis from Business Insider puts the percentage of people under 30 who use Facebook at 84 percent; the number above 65 who do is at 45 percent.
Mitt Romney, too, had a Facebook app in 2012, but it was less popular, perhaps for that reason. Despite this year's contested Senate races, the National Republican Senatorial Committee didn't bother making a new version of the app. "[W]e just didn't see the engagement numbers we needed to see to indicate that it was going to be transformative this cycle," an anonymous NRSC official told Yahoo.
But there's a whole different dimension in which campaigns use Facebook — a dimension that Facebook would much, much rather they utilize: Advertising. Facebook's Government and Politics page occasionally trumpets some of its success stories, pointing to things like Virginia Gov. Terry McAuliffe's (D) use of Facebook for his election in 2013. That summary of the effectiveness of Facebook ads is hosted, naturally, on Facebook's "Business" site.
It's not clear how effective those ads are in swinging elections. Shortly after this year's elections, Politico detailed an innovation from the campaign of Virginia Senate candidate Ed Gillespie (R). Seeing that conservative Facebook users liked the chain restaurant Buffalo Wild Wings, Gillespie went to one and had his picture taken, then spent money to promote the photo on users' walls. It was "a big-league bang for only a few bucks," Politico's Darren Samuelsohn wrote. But Gillespie lost. (McAuliffe won, which is probably why Facebook felt comfortable using him as a case study. See also: Barack Obama.)
There is at least one study that suggests that Facebook advertising may have helped influence an election. Journalist Simon Owens pointed to a 2011 Facebook post that outlined an evaluation of ads in a Florida ballot measure in 2010. The group Vote NO on 8 bought Facebook ads in Dade and Broward Counties to argue against the proposition, which then failed. Not only was there a big difference in the vote in the two counties where ads ran (19 percent more opposition), but people who were exposed to more online advertising voted 17 points against the proposition than those who saw fewer spots. Owens notes the results from a poll taken after the fact: "heavy web users who were on Facebook were 10 points more likely to vote no on 8 than Democrats (who may or may not have seen the ads) were."
We already knew that Facebook could drive people to the polls. In 2010, its experiment with an "I Voted!" button increased turnout by 340,000 during that year's midterms. In 2012, a different experiment ensured a select group of users saw more hard news as Election Day approached. The group that saw more hard news apparently turned out 3 percentage points more heavily.
Those studies prompted a bit of premature handwringing about Facebook's ability to swing election results. It's possible that Facebook could someday sell its ability to turn out voters, but it's not clear that this is in its immediate future. The company certainly doesn't mind the impression that it is a political kingmaker, it would just rather campaigns buy ads than make apps (particularly when the apps rile up privacy groups). Its Obama 2012 "Success Story" makes that clear. The app is barely mentioned. The amount raised from Facebook ads is front-and-center.
Update: This post has been updated to note when Facebook first announced its policy change on apps.