Cover Story

Mining for votes

With polls showing a Democratic lead, Republicans are using technology to target prospective supporters ever more precisely

Issue: "Demsnami," Nov. 4, 2006

MICHIGAN- When Susie Q. Voter leaves Wal-Mart pushing a shopping cart loaded with Dr. Pepper, Ozarka water, and Coors beer toward her Ford Explorer, she is likely to vote Republican. When Jane Doe heads to her Volvo from Whole Foods with a cartload of Pepsi, Evian water, and Budweiser, she will probably vote Democratic.

Farfetched? Maybe, but that's what GOP political consultants who prize data mining, the sophisticated connecting of disparate bits of information, predict-and with pundits and pollsters forecasting a Democratic tsunami, Republicans hope that data mining will give them a 4 percent to 5 percent election bump on Nov. 7. The GOP plans to "microtarget" Susie with advertising, phone calls, and even neighborly pressure, hoping that she will go to the polls-and to ignore Jane, hoping she will stay home.

Matthew Dowd, the Republican co-author of Applebee's America (Simon & Schuster, 2006), says flatly that "lifestyle is what determines political choice." Forget discussions of abstract political principles: "I would rather know where [voters] shop, what they buy, what kind of car they drive, what sports they watch, where the kids go to school." And forget economic class: "Income is no predictor" of voting behavior.

We see you’ve been enjoying the content on our exclusive member website. Ready to get unlimited access to all of WORLD’s member content?
Get your risk-free, 30-Day FREE Trial Membership right now.
(Don’t worry. It only takes a sec—and you don’t have to give us payment information right now.)

Get your risk-free, 30-Day FREE Trial Membership right now.

Data mining has long been used by businesses to target likely consumers for their products. But since 2000, Republicans have increasingly used a combination of public information and commercially available data about voters' magazine and media viewing habits, hobbies, and consumer preferences, in order to target particular voters with tailor-made messages. Democrats have been scrambling to catch up.

Dowd and Applebee's America co-authors Ron Fournier and Douglas Sosnik explain how a crucial insight after the 2000 election led to a radical shift in the way Republicans organized their campaigns. In analyzing election returns Dowd realized that it was senseless for campaigns to keep focusing on independent voters, a shrinking part of the electorate. Instead, successful campaigns should target their own "passive and inactive" voters via direct mail, cable TV, and the use of "navigators," neighborhood leaders with credibility among target groups.

Before the 2004 presidential election, data mining still had a reputation as "hocus pocus." As late as April 2004, New York Times reporter Joyce Purnick was skeptically writing, "Oh, to find that one piece of esoterica, or more realistically, that pattern of likes and dislikes, that will predict the politics of millions of voters. A belief in the wizardry of computers, in the idea that they can 'know' enough about everyone in the land to predict each voter's politics, is fueling experiments, hopes, dreams and even many an expert's election-year income in this campaign season."

But in the "battleground states" of 2004 Republicans used data mining to break down voters into 34 subgroups based on their buying habits, family makeup, and interests. Strategists conducted surveys with a subset of these voters to find out which issues made them angry and most motivated them to vote. With that information they designed specific campaign messages to reach each type. Bush campaign manager Ken Mehlman crowed after the election, "We were able to develop an exact kind of consumer model that corporate America does every day to predict how people vote-not based on where they live but how they live."

Republican success in 2004 surprised experts and put pressure on Democrats to develop similar programs. In Virginia's 2005 governor's race, Democrat Tim Kaine won in part by locating and micro-targeting likely Democratic voters in traditionally Republican parts of the state. In Texas, Democrat Donna Howard ran for a state House seat that came open in affluent west Austin and handily defeated her Republican opponent by using data mining to target voters most receptive to her appeals.

Overall, though, the national Democratic Party has lagged behind the GOP in micro-targeting, and now some political entrepreneurs are going their own way. Over the past year, for example, Hillary Clinton advisor Harold Ickes has pushed forward his own for-profit data-mining company, Data Warehouse, with funding coming largely from George Soros.

Both parties are seeing that many voters are less interested in platforms than in their sense of gut-level connection with candidates who they think display values such as strength, honesty, or compassion. Campaigns that need to "connect" emotionally with voters look to companies that have been able to connect with consumers-and that's the premise of Applebee's America. The 1,700-restaurant chain regularly conducts surveys to "understand the customer well enough that we can get an expanded number of them to vote for us with their wallets," as one Applebee's executive put it.


You must be a WORLD member to post comments.

    Keep Reading


    Power campaigns

    The GOP is fighting to maintain control of Congress…


    Troubling ties

    Under the Clinton State Department, influence from big money…