Skip to content
Link copied to clipboard

The Pulse: Complex science, simple goal: Votes

Politics is still a numbers game, but an ever more sophisticated one.

Target wrote the book on targeting.

The retail giant utilizes what's called "predictive analytics" to influence the purchasing habits of its customers. And now, that same level of sophistication is shaping our political campaigns as they seek to drive their vote to the polls Nov. 6.

Two books tell the story when read in tandem. The first is Charles Duhigg's The Power of Habit: Why We Do What We Do in Life and Business, which reveals how Target assigns its clients a code, referred to internally as a Guest ID, which tracks what we buy. What data does Target collect? Age. Marital status. Home address and commute time to the store. Credit card preference. And Web choices, to name a few.

This can be powerful information if used together with other details that are available for purchase, including "ethnicity, job history, the magazines you read, if you've ever declared bankruptcy or gotten a divorce, the year you bought (or lost) your house, where you went to college, what kinds of topics you talk about online, whether you prefer certain brands of coffee, paper towels, cereal or applesauce," Duhigg wrote for the New York Times Magazine earlier this year.

"In the last decade or so, our understanding of the science of habit formation has been completely transformed," he told me in a subsequent interview in March.

Here's a practical application: Duhigg says Target can sometimes determine when a customer is expecting a baby even if the woman never disclosed that information to the retailer, much less her family. How? By tracking purchases of unscented lotion, vitamins such as magnesium and zinc, washcloths, and cotton balls. The arrival of a newborn is one of the rare instances when purchasing habits are up for grabs and, if those patterns can be influenced, it becomes very profitable because, Duhigg told me, most of our "decisions . . . are actually habits."

When I said data mining could have great applicability to the political process, he said: "It's funny you should mention that." Duhigg told me that the Obama campaign had hired a chief scientist who comes from the habit-formation world. Hmmm.

Which brings me to Sasha Issenberg and his new book, Victory Lab: The Secret Science of Winning Campaigns, or as Politico calls it, "the Moneyball of Politics."

Gone are the days when getting out the vote simply involved party committee people working with street lists of registered voters whom they sought to mobilize strictly based on their party registration, says the former writer for Philadelphia Magazine and the Boston Globe.

"Now databases have gotten so good in politics that every time a campaign is contacting you - every time a committee person or ward leader or a volunteer is knocking on a door asking you, do you plan to vote, who do you plan to vote for, and what issues are important to you - that information is not getting thrown out on a clipboard at the end of the year but is sticking around. Now there are thousands of data points on each voter.

"You can run these complex statistical models, basically algorithms, that will give campaigns the confidence that they can predict . . . to a percentage probability - how likely you are to vote, who you're going to vote for, and what issues you're likely to care about."

Change has come, Issenberg says, because "all of a sudden people in politics realized, 'Wait, we can go and measure what we're doing - we can disentangle cause and effect.' "

Consider that, in 2005, Michigan political consultant Mark Grebner collaborated with two political scientists from Yale who were interested in finding new ways to motivate people to vote. One experiment involved sending voters a copy of their recent voting record and those of their neighbors, along with a promise to repeat the disclosure (all based on publicly available information) after the election. That threat increased voter turnout by 20 percent.

So which side benefits from the new sophistication?

"The left broadly is way ahead of where the right is right now," Issenberg said. "It's not a money issue. I think we underestimate how presidential reelection campaigns are so vastly different from a normal campaign."

Pay attention in the next month and you might see the net effect.

"I would look at the mail you get and try to deduce why somebody is sending it to you and think about if their assumptions seem right," Issenberg says. "Campaigns are sending you mail because they either think you're very likely to vote, so you're a regular voter, but that you're persuadable - or that you're someone who supports them and they're trying to get out the vote because you don't normally vote.

"If you're someone who votes often and you're a hard Obama supporter or a hard Romney supporter and either campaign is sending you mail, they're doing it wrong, and that probably means there's bad data or bad analysis underneath it."

No wonder they call it political "science."