Foursquare and 2010 Political Data

Could Foursquare predict voter sentiment and election results? A few online projects are testing how the location-based social platform can be utilized for election and political coverage.

  • Facebook
  • LinkedIn
  • Twitter
  • Email
  • linkText
In 2008, voters went to Twitter to announce where and how they voted, and to read other voters' reactions. In 2010, the social media platform to watch this Election Day may be Foursquare. The location-based platform wasn't established until last year, so this year's elections provide the first opportunities to experiment with how it can be used for political purposes.

One Foursquare experiment comes by way of Toronto. On Oct. 25, Torontonians voted for their new municipal leaders, including mayor. Local newspaper National Post, which uses Foursquare to connect its news stories to places around the city, set up a Foursquare location for voters (Toronto Election 2010) and encouraged them to "check-in", share who they voted for and why. Staffers hoped to see if Foursquare could be an alternative or extension of traditional exit polling. Over 320 people checked in at 'Toronto Election 2010' and 18 shared their votes.

Did the National Post's Foursquare exit poll predict the election? Well, no. Users who shared who they voted for tended to vote for mayoral candidate George Smitherman, who lost the election to another candidate, Rob Ford.

"It's tempting to speculate that people with liberal views (i.e. Smitherman supporters) are more likely to be early adopters on emerging platforms such as Foursquare than their conservative, Ford-supporting counterparts," says National Post Senior Producer Chris Boutet in an e-mail. "But I think we can mostly chalk this up to the sample size just being too small to provide an accurate reflection on how the vote would go." The Foursquare sample is quite small when you consider that 52.6 percent of the city's 1.5 million residents came out to vote.

Boutet also noticed a possibility that Foursquare users were hesitant to share who they voted for, so on election day he set up an online poll that he promoted off the Foursquare event. That poll received over 2,000 votes and was more accurate in predicting the election.

Would Boutet try another election-related project on Foursquare again? He says the paper "would definitely" do something similar for future elections, but would tweak the execution of it to maybe include multiple polling locations to see if that would increase visibility and the number of check-ins.

Like Torontonians, Americans can also check in at the polls when they go to vote Nov. 2. Foursquare partnered with a number of election-focused organizations to bring "I Voted," a data visualization project that will use the number of check-ins at over 100,000 polling places to show how big voter turnout is at different locations. "With over four million users, Foursquare is now at the scale where check-ins communicate a larger trend and we're excited to make this data more accessible to the public," says Foursquare's Eric Friedman in a statement.

In addition, the project will bring the "I Voted" sticker into the digital age. Users can earn an "I Voted" badge if they check in at a polling location, which will be posted on the user's Foursquare profile.



The Foursquare political experiments won't end after Election Day. GeoPollster, a feature that launched in early October, will continue to track Foursquare users' political sentiments in real time. The way it works is that when users opt in to GeoPollster, they are asked to affiliate themselves with the party they would vote for if the election was right now: Democratic, Republican, Green, Libertarian or Constitutionalist. This information is hidden from other users.

If someone checks in at, say, a coffee shop or the local DMV, GeoPollster counts that check in as a vote for his or her chosen party. People who check in most often at single venues (known as "mayors") have their vote counted slightly more than a visitor, which helps preserve the game aspect of Foursquare and also gives priority to local user sentiment. The number and strength of party-affiliated check-ins determines which party is in power at different venues, cities and states. By GeoPollster's count, Democrats have control of the Kennedy Space Center in Florida, and Republicans have control of Fulton County Superior Court in Atlanta.

GeoPollster's map of party control in each state looks much more colorful and divergent than the red and blue maps voters recognize. California, Colorado and Nevada are ruled by Green Party-affiliated users, whereas the Carolinas, Louisiana and Texas are ruled by Libertarian users. "I think GeoPollster allows people to vote for the true party they support and not what's just presented on the ballot," says GeoPollster co-founder Adam Kraft.

Users can and are encouraged to change their party affiliation at any time. Kraft has only seen one person (out of 272 current users) change his party affiliation, but he hopes that GeoPollster will be able to predict trends based on world events and party switches. Kraft would also like to eventually ask its users single-issue questions to see where the user base stands.

What will the Foursquare turnout and response be on Nov. 2 and beyond? Who knows. But perhaps some of the political projects on Foursquare in 2010 could be the state of a new way to track voter sentiment in upcoming mayoral races (Feb. 22 in Chicago, anyone?), 2012, and beyond.

  • Facebook
  • LinkedIn
  • Twitter
  • Email
  • linkText
Tina Trenkner is the Deputy Editor for GOVERNING.com. She edits the Technology and Health newsletters.
Special Projects