Political data, once the reserve of presidential campaigns, is spreading to local races


A visualization of likely gun owners, according to Clarity Campaigns. (Clarity Campaigns)

The November midterms aren't the only race Republicans and Democrats are running this year.

As Election Day approaches, both parties are rushing to put new political technology into the hands of smaller campaigns that may lack the resources or staff of a major presidential effort. Candidates competing for seats in state and local races are increasingly gaining access to sophisticated databases and predictive modeling tools, much like those that earned President Obama so much attention in 2012. As a result, the 2014 election season is likely to see a much wider use of data to target likely voters, organize volunteers and consolidate voter lists.

The GOP's new toolbox

The Republican National Committee said Wednesday that it's trying to persuade conservative campaigns to adopt a suite of free tools it has developed in recent months to give candidates an edge. The package includes a massive voter file known as OneData, which contains information on 190 million active U.S. voters gathered from all 50 states. Republican campaigns will have access to information on another 63 million Americans through data gathered from commercial sources such as Axiom and Experian, the committee said. Altogether, the data tracks nearly two decades of Americans' voting history, including whether they own hunting and fishing licenses, that could indicate their positions on political issues.

The RNC is also pitching small campaigns a tool called RNC Foresight, which uses predictive analytics to determine whether a voter will vote Republican. (Predictive analytics is the practice of matching a given voter against aggregate demographic profiles to understand how that voter may behave on Election Day.) That data could help campaigns determine how to allocate resources and voter outreach efforts.

The GOP will be educating state party officials and campaigns about the new tools in its spring meeting this week, the RNC said. The effort begins with congressional, gubernatorial and mayoral races. But the committee hopes those running for lower offices, such as county commissioner, will adopt the new resources, too.

"We can bring this out not just in one presidential campaign but every Republican campaign," said an RNC official involved in developing the tools, who spoke on condition of anonymity because the briefings were private.

Republicans have been testing parts of the suite in places such as Massachusetts and New Jersey. It was deployed for the first time in full during the March special congressional election in Florida, where David Jolly defeated Democrat Alex Sink by a few thousand votes in what the RNC said was a "live-fire test" of the new technology. But now the RNC is offering the tools to all conservative campaigns in an effort to correct what critics, including those within the party, have called a deficit in technology compared to Democrats.

A model in New Jersey

The use of data and analytics is spreading to downticket Democratic candidates, too. In New Jersey's gubernatorial election last year, Republican Gov. Chris Christie swept the ballot, winning more than 60 percent of the popular vote. But Democrats in Middlesex County thought that they could use computer models to target voters in smaller races.

It was an untested tactic. Over resistance from some party members, Democratic strategist Kevin McCabe tried to identify and persuade what he called "the right voters," rather than focusing on a broad group of existing supporters.

"Identifying the right voter was instrumental to our success at a county level," said McCabe in an interview. "People had told me, 'You've got to get out the base, you've got to get out the base.' And we do. But we've got to make sure we get the right voters. That was difficult for some to grasp."

McCabe hired Clarity Campaigns, a political consulting firm, to design a predictive model that approximated the kind of voter he was looking for. The model advised Democrats not to reach out to some voters, which made some uneasy, said Clarity co-founder Tom Bonier. But what the campaigns gave up in breadth was made up in depth.

"They were able to use the models to not only target the right voters, but also target messages," said Bonier, "sending pieces of mail on gun control to households that the model identified as likely to move on the issue, while talking about taxes to other households."

Democrats went on to win all four countywide races at stake.

Earlier this year, the Democratic National Committee announced Project Ivy, its own effort to apply analytics and data-driven decision-making to non-presidential contests. "The key point about Project Ivy is that it's about continuing to push the envelope in terms of trying to push the toolset to smaller campaigns," said DNC tech director Andrew Brown. "The investments we've made are not 'will be' or 'could be' but are being utilized downballot in a smart way."

These days, all political data are local.

Brian Fung covers technology for The Washington Post, focusing on telecom, broadband and digital politics. Before joining the Post, he was the technology correspondent for National Journal and an associate editor at the Atlantic.
Comments
Show Comments
Most Read Business
Next Story
Andrea Peterson · May 7