You are not logged in. LOG IN NOW >

Where Obama's Ground Soldiers Were, and Who They Are

BY Nick Judd | Thursday, November 8 2012

Ryan Enos / Eitan Hersh

A just-concluded research project studied thousands of Obama for America ground volunteers as they knocked on doors and made phone calls in their effort to get Barack Obama into the White House for another four years. Field organizers and volunteers answered a survey presented to them as they accessed a common piece of campaign software. This map shows where they were as they were answering the survey — campaign data exhaust that proves Obama's ground game was at its strongest exactly where Mitt Romney's campaign needed it to be weak.

According to this map, the Obama campaign concentrated its efforts on the ground game in Colorado, Pennsylvania, Ohio, New Hampshire, Iowa, Virginia, and North Carolina. Another graph not adjusted to take into account electoral votes shows Wisconsin and Michigan in the mix, too.

Much has been said already this year about how political campaigns, especially Democrats and especially the 2012 campaign to re-elect Barack Obama, have been furiously working to understand the minds of American voters with empirical precision.

But academic researchers are also hoping to find out more about Obama's campaigners themselves.

One result of the Obama campaign's willingness to cooperate with academic researchers is the above map, which actually doesn't say anything about American voters. Instead, it's a map of how many Obama for America field workers were in each state relative to that state's importance in the race to 270 electoral votes. It was released today as one of the first results of academic research into the mind of the Obama volunteer. Because the researchers were at work for months, they were even able to track where volunteers were over time.

From June 11 to Nov. 6, they worked with the cooperation of Obama for America and the software company NGP VAN to present questionnaires at random to volunteers and campaign workers on Obama's re-election effort. The result, they announced today, is a dataset of survey responses from nearly 4,000 Obama for America campaign workers selected at random from across the country.

Volunteers and field workers on Obama's campaign used NGP VAN's VoteBuilder software to record their every contact with voters. It is the same tool that many Democratic campaigns also use. The two researchers used VoteBuilder to open a window into the Obama campaign's inner workings.

Workers logging in to NGP-VAN software had a one in 100 chance of being asked to participate in the survey, Ryan D. Enos, a Harvard University assistant professor of government and one of the two principal researchers on this project, told me today. They were asked a series of questions to track how they thought the race was going, among other things, including questions designed to mirror what pollsters were asking American voters in mass phone surveys at the same time.

"We don't really know a lot about this army of people that works on campaigns," said Eitan Hersh, a Yale University assistant professor and the other half of the research duo. "We know they're out there. We know they're knocking on doors. We have a lot of stereotypes, that they're students who are taking time off from school, that they're retirees, that they're union members, but we don't know who these people are."

Consensus is building that the future of American politics is somewhat antediluvian, not in a torrent of expensive direct mail or television advertising but in a street-level organizing based on coalitions and a lot of one-to-one contact. It takes a certain class of campaigner to wage that kind of campaign, with training and understanding Democrats have in spades and Republicans willingly admit they will be trying for the first time.

Who is going to inherit American politics? This survey might have something to say about exactly that question, at least for Democrats.

If Republicans have a lesser understanding of themselves headed into whatever succeeds the days of television advertising and direct mail — days which, I think, people may soon be ready to say are over — it's mostly their own fault.

"It's much easier, in general, to work with the Democratic side, at least from the perspective of an academic," Hersh told me today.

There are two reasons, he said: One, the Republican operation is much more centralized.

"The Democratic operation has a much more open-source flavor to it," Hersh said. "They give logins to their data to a lot of people and their view is that this helps, that more people working on this data the better the outcomes will be. On the Republican side, you get the sense that it's more centralized, that they're more into controlling the data, and I think that is part of the problem in terms of our access for academics."

The other part, Hersh said, is the presumption that any academic is likely to be a liberal and as such is not to be trusted. Earlier in our conversation, he also expressed concern that if he tried to work with Republicans and Democrats in the same cycle, he wouldn't be able to work with either.

"They didn't explicitly say don't work on this with the Republicans," he said, "but everything in the political world is partisan, and just like a mail firm that just does direct mail doesn't work with both sides, it's hard to build a level of trust with any campaign if you're working with the other side. I think that's too bad, but I think that's the price you pay for studying sitting campaigns in a very in-depth way."

This post has been updated.