Why Campaigns Are Happy Your Vote Isn't as Private as Many Think It Is
BY Nick Judd | Monday, October 22 2012
The mailer that arrived in Ann Althouse's mailbox in June really creeped her out.
"Why do so many people fail to vote?" it asked. "We've been talking about the problem for years, but it only seems to get worse. This year, we're taking a new approach. We're sending this mailing to you and your neighbors to publicize who does and does not vote."
The mailer was the work of the Greater Wisconsin Political Fund, a progressive group hoping to drive turnout ahead of this summer's critical recall elections. On it was Althouse's name as well as those of several of her neighbors, along with whether they voted in the 2008 and 2010 general elections. Althouse, a conservative law professor and blogger, was shocked — she called the practice "truly despicable," complaining that "your vote is private" and "this is an effort to shame and pressure people about voting" — but by the time this strategy reached Wisconsin, it was already old hat.
This tactic, called "social pressure" by some and "shaming" by others, is partly driving the latest round of hand-wringing over the data-driven, highly targeted political present. Rather than using commercial marketing data about voters, these strategies use public records and social connections to convince people to vote.
There are shades of this in the tools now in play from both presidential campaigns. The latest round of fundraising emails from the Obama campaign includes messages that remind supporters of how much they have given to the campaign so far — or calls them out for not giving anything. Another effort asks supporters who have signed in to the Obama for America Facebook app, thus sharing their friends lists with the campaign, to share a get-out-the-vote message with acquaintances in swing states. Mitt Romney's campaign just launched a similar tool. In a recent article in The New York Times, Charles Duhigg, author of a book that addresses targeted marketing, quotes an unnamed Democratic consultant wondering aloud if "this is the year to start shaming" — but it seems the finger-wagging and elbow-nudging has already begun. These are just parts of an interconnected web of experiments, partly taking place through mailers, partly driving new technology-driven communication over social networks. These new initiatives play on the concept that what voters know about one another may turn out to be more important than what the campaigns themselves have on file.
What they know, and when they knew it
"There is some evidence, some recent evidence that suggests that voters are actually much less responsive to microtargeted or narrowly targeted messages than they are to more broad-based appeals," said Costas Panagopoulos, an associate professor of political science and director of the Center for Electoral Politics at Fordham University.
On the other hand, he said, several studies suggest that voters respond strongly to being contacted with their own personal information.
"So for example when you disclose people's recent voting history to others, that's something that voters are very responsive to. And in the case of prior voting the academic literature has shown that voters are more likely to participate in elections when you disclose or threaten to disclose their electoral participation, presumably because voting is a social norm or a socially desirable activity and they don't want to be perceived as violating those norms."
A landmark 2008 study led by Yale University professor Alan Gerber found so-called "social pressure" of this nature, threatening to out people to their neighbors as civic slackers, could increase voter turnout rates by upwards of eight percent more than a mailer that merely asked the recipient to go vote without mentioning personal information. As soon as the news got out that "social pressure" works, campaigns began testing the practice and academics continued to study it in the 2008 and 2010 election cycles.
Panagopoulos says the rate at which people objected to being targeted like this in an academic study was relatively low. But when the Greater Wisconsin Political Fund gave this a shot, Althouse wasn't the only one to think it was a violation of privacy. Plenty of other people also objected to this use of voter data. Supporters of an effort to recall a Republican state senator derided the idea on a public Facebook thread and in the forums of Democratic Underground, a progressive online community, expressed the same discomfort with this use of personal information. The volume of protest became so great that on June 4, three days after Althouse received her copy of the mailer, the Wisconsin Government Accountability Board put out a press release pointing out that the voter rolls were public, the mailers were legal, and none of the blame could be placed at the feet of Wisconsin officials.
This wasn't the only instance in which "voter shaming" generated more anger than embarrassment. In 2010, Senate candidate Mike Lee sent an email to supporters in Utah with an attachment that included the names and contact information of voters who had voted in presidential elections, but not midterms. These people, the email suggested, needed to be educated about the importance of midterm elections. The Daily Herald of Provo, Utah, quotes a local voter, Andy Gibbons, as calling the email "despicable."
Panagopoulos, the Fordham associate professor, is an author of a study that experimented with what would happen if the message was less heavy-handed. Rather than trying to shame or pressure voters, his study focused on messages thanking people for voting in the past. While the results weren't as dramatic as messages that sought to motivate by fear, they still correlated with a notable rise in turnout.
"Most people don't react with hostility," Panagopoulos said, "but some voters find it offensive, and I think these are the people who have concerns about how much information is publicly available about their political behavior."
As Shane Hamlin, co-director of elections in Washington State, explains, how you vote is secret, and nobody is suggesting that it shouldn't be. But when you did or did not vote has always been a matter of public record.
In June, labor groups used a tool called Amicus during the Wisconsin recall effort. Amicus users check their Facebook friends against a voter file, and come back with lists of their friends who a given campaign or political group believes might be persuaded to go out and vote. In an interview, Amicus co-founder Seth Bannon explained that this is motivated in part by the same "social pressure" theories that informed the Wisconsin mailers. He cited the Alan Gerber study by name. Another recent study also found that get-out-the-vote messages on Facebook, pressure or no, is correlated with higher turnout.
"In the old days a campaign or political party could send a poll watcher to the sites to see who voted and who didn't and take that information back to the headquarters and get on the phone to remind people to vote," Hamlin said.
"That's the way it happened 10 years ago," Hamlin continued. "The availability of the Facebook aspect isn't really much different except that it's instant. I think it's still an important part of the democratic process. Voting is done in private and secret, but it's still a community experience."
Personal Democracy Media is also involved in an initiative that puts this idea to use. Supported by the Ford Foundation, TechPresident's parent company is collaborating with the Internet freedom advocacy group Center for Rights on Vote with Friends. This tool shows you which of your friends appear in a database as being registered to vote and gives you the opportunity to nudge them towards voting. Like all voter records, the database isn't perfect — for example, it tried and failed to find my own voter registration.
Inaccurate data is just one of many privacy concerns stirred up when it becomes obvious that voter registration information is being used to figure out who to target. After the historically close 2004 governor's race in Washington, the state legislature there made changes to election law in the hopes of making it less likely for a close race to be decided as much by rounding errors as on who cast a ballot. One of those changes included making each voter's date of birth available for public inspection, a change Hamlin, of the Washington Secretary of State, says his office fought against and lost. When Washington's Help America Vote Act-mandated statewide voter database went online, each voter's data of birth came online with it.
"You get a statewide list from us for the first time in 2006 and then some bloggers, who were still convinced that the 2004 election didn't come out the way it should have, created searchable databases," Hamlin explained to me. "And then they were discovering all this information about themselves that was public that they did not know was public."
That, he said, generated a lot of phone calls. What's happening now, he suggested, is similar: In an election year, people are confronted with more evidence that there's a difference between what they think is public about them and what actually is.
Campaigns and researchers alike know that at least some people are unsettled by increasing access to information about voting history. Christopher Mann, an assistant professor at the University of Miami, conducted research that focused specifically on voter backlash.
"You can deliver that same information of, 'we know,' but deliver it in a way that is less threatening and Big Brotherish," Mann said. "It can be delivered, in my research it was delivered in the sense of being helpful. 'We've seen that you don't vote in every election. Here are some resources to help you vote.'"
But wrapping that "we know" message in the robes of academia is not always enough to cool people down. Also this summer, a Harvard University Ph.D student in economics and a researcher at Universidad Nacional de La Plata, Argentina, led a study that sent mailers at random to people listed in the Federal Elections Commission database as donors who had given $200 or more. Those mailers listed each donor's contribution along with the contributions of a few of their neighbors, identified by first name and last initial. The Chronicle of Higher Education quotes Orlando, Fla. resident Susan Kelley, who received a mailer:
“While I am well aware that the information is public, as is a great deal of other information about each of us, I was offended that Harvard University would feel it was in the interests of research to urge people to view each other’s political contributions,” she writes in an e-mail. “I am no more interested in the contributions of my neighbors to their favorite politicians than I am in how they vote.”
When contacted by email, the researcher on the Harvard project, Perez Truglia, declined to go into details about his study, saying data analysis was not yet complete.
"One of our aims is to study the implications of the public’s awareness about the open nature of campaign contributions," he wrote in an email. "We hope that the research will shed light on the advantages and disadvantages of different disclosure policies, which we believe is a very important issue."
A more open future?
For people like Seth Bannon, Amicus' founder, a future where everyone's political cards are on the table is inevitable. Bannon, like Facebook's Mark Zuckerberg, advocates for transparency as a social force while at the same time the success of his business depends on it. Nonetheless, that's the future he sees: One where people know and accept that attendance is taken in the public square and their friends are keeping track.
"We're definitely moving towards a world where my friends will know whether I voted in the last election or not and they're going to let me hear about it if I didn't," Bannon said. "I think public perception hasn't caught up to the technology yet, but it quickly is."
Some voters, like the Wisconsin law professor Ann Althouse, say that not voting is a legitimate choice. It's as much a way to participate as voting, and people shouldn't be publicly called out for deciding not to vote. Bannon is not one of those people.
"I think that voting is a civic responsibility as citizens," he said, "and you're not going to change the process by not participating in it."
Just more noise on the channel
Political campaigns know that this type of contact can be disconcerting, said Mann, the University of Miami professor who did research on the backlash potential of social pressure to vote. They also know that studies indicate it can help them win.
What isn't clear is if any kind of communications — targeted advertising, overt pressure using personal information, anything — can cut through the intense amount of noise directed at voters in key states like Ohio.
"I talk to a lot of campaign professionals," Mann told me. "What's interesting is there was some reluctance to at first use any social pressure and then they got comfortable using some sort of social pressure. And as the stakes go up, campaigns start to think more about using tactics they might not otherwise."
"Maybe they thought that doing shaming might have felt over the top or inappropriate in a past election but now it feels justified," he said later on in our conversation. "What's not clear to me is if it's going to make a difference given all the other things that are out there."
Ohio is awash in advertising money from the campaigns and outside groups. The state has already been open for early voting for two weeks, and has another two weeks left to go. The Romney and Obama campaigns are treating every day in Ohio like it's election day, Mann said. That barrage is making Ohioans tune out.
"Unrelated to social pressure, I'm conducting experiments in battlefield states," Mann said. "You can't get people to answer the phone in Ohio."
That's where these emerging experiments at the margins, focused on person-to-person contact, begin to make sense. The Obama campaign, for instance, has in recent days been sending emails to supporters asking them to contact friends in battleground states with early voting. Obama supporters who have signed in to the Obama for America Facebook app have shared their friends list with the campaign. Some of them are getting emails asking them to share Facebook messages with friends in states like Ohio and Nevada that have the potential to decide the entire election. The Romney campaign has also launched an app that rifles through users' Facebook friends and highlights potential voters to contact on the campaign's behalf. Neither of those appear to include a Big-Brother style message about following up to tell everyone if the voter cast a ballot, but they do ask supporters to reach out based on things the campaigns know about those supporters' friends, like where they live.
"The logic behind this is the impact of these messages will be stronger from their friends," Mann said. "That's from psychology. There's a common-sense piece here that a big challenge for campaigns in this electoral environment is delivering a message that they pay attention to. It isn't the fourth phone call they receive in a night, it isn't another piece in a stack of mail."
Both presidential campaigns now seem to be trying to find a path to victory that plays on this combination of available information, available connections, and social pressure, but does so in a way that is subtle enough not to piss anyone off. I asked Ann Althouse, the Wisconsin law professor who protested against getting that overtly threatening mailer from the Greater Wisconsin Political Fund, about the argument that a voter's history is open and voting is a civic duty. I also pointed out to her that new tools allowed campaigns to identify the likeliest voters from among a supporter's Facebook friends and give that supporter ideas about how to gently nudge them towards the ballot box.
It's one thing to keep track of who voted to make sure an election is fair and the tally is accurate, she said, and quite another to use it for a political purpose.
"I'm aware of the research that says this social pressure works," Althouse wrote to me in an email, "but if using it subjects you to condemnation, it can backfire. I hope it backfires. I think it's creepy, ugly, and embarrassing for the candidate who stoops to this method. I feel the same way about playing on racial fears. You're free to do that, and you might sway some votes that way, there will be a backlash of condemnation that makes it a net loss."
What about friends who reach out on Facebook, I asked, or make a phone call, that's obviously informed by a campaign's hope that it will pressure her to vote?
"If such a call went through at my house, the caller -- if it was someone I knew so I answered the phone -- would get a piece of my mind," she wrote, "and would not get off the phone with me until he exhibited shame and apologized for calling."
This post has been updated. Charles Duhigg's book, "The Power of Habit," is focused on studies of behavior and how they are applied to business, not only targeted marketing.
This post has been corrected to give Shane Hamlin's proper title. He is co-director of elections, not assistant secretary of state.
This post has been corrected to properly explain mailers sent out this summer with donor information. They included donors' first names and last initials, not first initial and last name.