Everybody thought it was a little strange when Barack Obama made a campaign stop at Reddit back in August to host an Ask Me Anything session. Why would he be interested in fielding off the wall inquiries from the Internet’s nerd kings? How would it be better to spend an afternoon staring at a laptop screen instead of going out into the streets and kissing some babies? “Why did we put Barack Obama on Reddit?" one Obama campaign official explained. "Because a whole bunch of our turnout targets were on Reddit.”
The 2012 election will go down in history as the year that data won. More specifically, it’s the year that brought out the science in political science, as an unprecedented amount of information about voters became available. This led to a new cottage industry in taking that information, modeling it for tens of thousands of different scenarios, and then planning efforts based on what would have the right impact in the right places. Simply put, the candidates who designed their campaigns to leverage the power of big data in their favor had a big advantage over those who did things 20th-century-style.
The Obama campaign is a great example. In 2008, Obama won his first term in the White House with a focus on small donations and a forward-looking attitude about the power of social media. They even hired Facebook founder Chris Hughes to run those social media efforts. At the time, everybody was just gushing about what an innovative campaign it was and how it would change the way American politicians ran their campaigns for decades to come. That prediction turned out to be all too true for Obama for America in 2012, because 2008 not only informed them of smart ways to do fund-raising or how to involve social media. It also gave them an unprecedented amount of voter data to play with, an opportunity that OFA was quick to seize.
Jim Messina, Obama’s campaign manager, made a point to put big data at the center of his strategy from the very beginning. “We are going to measure every single thing in this campaign,” Messina said, just before hiring a “chief scientist” to oversee an analytics team five times the size of the one in 2008. The team acted with such secrecy that even the rest of the staff in the Chicago campaign headquarters usually didn’t know what they were doing. Put simply, they were crunching numbers. Put honestly, they were combining all of the data collected in 2008 — from voting records to donation history to voter demographics to, well, pretty much anything else you can think of. Then they built models. Lots and lots of models that would inevitably help them predict what voters would do when and why.
“We could [predict] people who were going to give online. We could model people who were going to give through mail. We could model volunteers,” a senior OFA adviser told TIME in the last days of the campaign. “In the end, modeling became something way bigger for us in ’12 than in ’08 because it made our time more efficient.” It’s also how they planned their days. “We ran the election 66,000 times every night,” said another official. “And every morning we got the spit-out — here are your chances of winning these states. And that is how we allocated resources.”
It wasn’t just the campaigns that were leveraging big data, either. New York Times number wizard Nate Silver (pictured above) gained widespread praise for correctly predicting the outcome of 49 out of 50 states in the 2008 election with his knack for aggregating polling data, but he became a household name in 2012 well before election day. In the final weeks of the campaigns, a number of pundits expressed doubt about his methods. Silver wasn’t necessarily doing anything special or new. In fact, The Washington Post’s Ezra Klein called the Silver approach “little more than a sophisticated form of poll aggregation.”
What Silver made clear, though, was that the right treatment of the increasingly massive amount of polling data out there could produce mathematically accurate predictions. With Florida still too close to call, Silver is batting a thousand with his predictions for 2012. It’s also worth noting that Silver accounted for 20 percent of The New York Times’s web traffic just before the election.
So data steered the way to success for the Obama campaign and helped The New York Times become the most trustworthy source of election information out there. Is there anything that data can’t do? That remains to be seen. We call all expect everyone to double down on big data in 2016, and the Republicans are already talking up their plans to make it the first ever mobile first campaign. In the meantime, the numbers have spoken. Politics is more a numbers game than it ever was before.