If you voted this election season, President Obama almost certainly has a file on you. His vast campaign database includes information on voters’ magazine subscriptions, car registrations, housing values and hunting licenses, along with scores estimating how likely they were to cast ballots for his reelection.
And although the election is over, Obama’s database is just getting started.
Democrats are pressing to expand and redeploy the most sophisticated voter list in history, beginning with next year’s gubernatorial races in Virginia and New Jersey and extending to campaigns for years to come. The prospect already has some Republicans worried.
“It’s always hard to play catch-up,” said Peter Pasi, a Republican direct marketer who worked on Rick Santorum’s presidential primary campaign. “It can be done by 2016. I’m much more doubtful it can happen by 2014.”
The database consists of voting records and political donation histories bolstered by vast amounts of personal but publicly available consumer data, say campaign officials and others familiar with the operation. It could record hundreds of pieces of information for each voter.
Campaign workers added far more detail through a broad range of voter contacts — in person, on the phone, via e-mail or through visits to the campaign’s Web site. Those who used its Facebook app, for example, had their files updated with lists of their Facebook friends, along with scores measuring the intensity of those relationships and whether they lived in swing states. If their last names sounded Hispanic, a key target group for the campaign, the database recorded that, too.
The result was a digital operation far more elaborate than the one mounted by Obama’s Republican rival, Mitt Romney, who collected less data and deployed it less effectively, officials from both parties say.
To maintain their advantage, Democrats say they must navigate the inevitable intraparty squabbles over who gets access now that the unifying forces of a billion-dollar presidential campaign are gone.
“If this is all we do with this technology, I think it will be a wasted opportunity,” said Michael Slaby, the Obama campaign’s chief integration and innovation officer.
Tests of whether Obama’s database can be successfully redeployed will come even sooner. Terence R. McAuliffe, a party insider who ran unsuccessfully for governor of Virginia in 2009, has inquired about the data for his gubernatorial campaign next year, say those familiar with the conversations.
“We have been communicating to Obama for America all along about the importance of receiving that data, since Virginia has a 2013 election,” said Brian Moran, the outgoing chairman of the Virginia Democratic Party.
Although McAuliffe is the early Democratic front-runner, many in the party say individual candidates should receive access to such data only after winning the nomination — something that in Virginia can’t happen before the June primary, leaving only a few months before the November general election. The short time frame may make a full data set, should McAuliffe get it, even more valuable.
All Democratic candidates have access to the party’s lists, which include voting and donation histories along with some consumer data. What Obama’s database adds are the more fine-grained analyses of what issues matter most to voters and how best to motivate them to donate, volunteer and vote.
But there are serious logistical challenges to keeping updated a database as large and as detailed as Obama’s, which is why campaign officials are debating how to proceed even though there is wide agreement on the desire to help fellow Democrats and like-minded independent groups.
Slaby, the campaign official, said the database in the near term could be used to organize support for the president’s legislative agenda but eventually might go to the Democratic National Committee or Obama’s presidential library committee once it is established. Or, he said, it could go to a group created to nurture and deploy the database most effectively. No existing group has the technical resources to manage the data, he said.
Slaby said of Obama, “A lot of this will rest on him and what he wants his legacy and the legacy of this organization to mean.”
The database powered nearly everything about Obama’s campaign, including fundraising, identifying likely supporters and urging them to vote. This resulted in an operational edge that helped a candidate with a slim margin in the overall national vote to trounce Romney in the state-by-state electoral college contests.
Obama was able to collect and use personal data largely free of the restrictions that govern similar efforts by private companies. Neither the Federal Trade Commission, which has investigated the handling of personal data by Google, Facebook and other companies, nor the Federal Election Commission has jurisdiction over how campaigns use such information, officials at those agencies say.
Privacy advocates say the opportunity for abuse — by Obama, Romney or any other politician’s campaign — is serious, as is the danger of hackers stealing the data. Voters who willingly gave campaigns such information may not have understood that it would be passed on to the party or other candidates, even though disclosures on Web sites and Facebook apps warn of that possibility.
Chris Soghoian, an analyst at the American Civil Liberties Union and a former FTC technologist, said voters should worry that the interests of politicians and commercial data brokers have aligned, making legal restrictions of data collection less likely.
“They’re going to be loath to regulate those companies if they are relying on them to target voters,” he said.
Slaby said the campaign took great care with the data it collected and will ensure that whoever takes it over will protect it. Such efforts, though, take unusual resources, he said. Building the campaign’s technological systems took nearly two years and, at their peak, involved about 120 paid employees working with data provided by hundreds of thousands of volunteers.
Republicans once held the edge in using technology to identify and motivate voters. After Sen. John F. Kerry (D-Mass.) lost to President George W. Bush in 2004, Democrats invested in building better voter lists and developing a new generation of political operatives skilled in the science of persuasion and motivation.
Although Obama’s 2008 election was hailed for its technological advances, campaign officials acknowledge that the operation fell far short of its hype.
With the benefit of four years of lead time, the campaign was determined to make better use of increasingly sophisticated technology. Driving this was Obama’s data-minded campaign manager, Jim Messina. Among his mentors was Google Executive Chairman Eric Schmidt, who was a regular visitor to what many have said resembled an Internet start-up company within the Chicago campaign headquarters.
The campaign invested heavily in engineers and technologists, including many who had never worked in politics, and used Amazon Web Services to host the voter database on its cloud servers. The key was a program the campaign built — called Narwhal after a predatory whale whose single tusk makes it look a bit like a fat, finned unicorn — that consolidated lists of voters and donors, often collected over years by state party officials and campaigns.
Narwhal allowed related pieces of software, such as those used by field organizers and call center workers, to draw on the information in the voter database and continually update it.
Slaby and others from the campaign said that although it relied on detailed analyses of cable television viewing habits and Web traffic, personal information from those sources was made anonymous and did not flow back into the voter database.
The most important information, officials said, was provided by voters themselves whenever they had contact with the campaign, in person or online, enriching the database with e-mail addresses, cellphone numbers and, crucially, information about what issues most concerned them.
This allowed the campaign’s analysts to test the effectiveness of messages aimed at narrow demographic slices — single women in their 30s worried about health care, for example. Although it was often described as “micro-targeting,” Slaby said the most important element was what he called “micro-listening.”
“If people tell us they’re interested in cats, we probably took that down,” he said.