The Victory Lab (48 page)

Women’s Voices donors signed off on this new so-called Promise technique, and Gardner included it in the group’s plans for turnout operations in 2008. At the same time, Malchow giddily went to an Analyst Institute lunch with PowerPoint slides documenting the Kentucky test, hoping that his peers at other liberal groups would be as eager to put social pressure to work in their 2008 voter contact programs. The institute later promoted Promise as part of its best practices for improving turnout rates. But even the members who found the experiment fascinating also found it hard to imagine such manipulation finding a place in the political toolbox—they couldn’t visualize their name as the return address on a letter that told voters they were subjects in a study, even if it worked.

Malchow knew he had more work to do if he was to find a way of provoking anxiety in people for not voting without antagonizing them. “This is the frontier—thinking about ways to do this that are unoffensive,” says Malchow. What if instead of embarrassing people for not casting ballots, he just sent a list of their neighbors who voted all the time? He mentioned this idea to Green, who referred Malchow to Costas Panagopoulos, a former Yale postdoctoral researcher who had collaborated on research with Green at the time of his Michigan experiment with Grebner.

Panagopoulos had already been investigating other forms of social surveillance, such as the increasing affection that law enforcement officials had for publicizing the names and images of sex offenders, johns, and those delinquent on child support payments. Along the way, he learned about two newspapers that had applied a similar logic to citizenship.
In 1994, the
Dallas Examiner
published local electoral rolls with an indication of who had voted in a recent election and who hadn’t.
Before the 2006 elections, the
Tennessee Tribune
ran its own list of nonvoters in selected Nashville city council districts. The papers, both targeted at local African-American communities, claimed that they had boosted turnout when they introduced the disclosure program in their pages.

Panagopoulos decided to put the method to the test. He identified three small midwestern towns that would be conducting nonpartisan municipal elections in November 2007 and randomly selected households in each to receive pre-election postcards. In Monticello, Iowa, and Holland, Michigan, the cards told recipients that a list of those who voted would appear in the local newspaper after the election. “The names of voters who did not vote will
not
be published because only voters deserve special recognition,” the cards read. In Ely, Iowa, postcards made the opposite threat: the local paper would publish a roster of deadbeats only. “The names of those who took the time to vote will not appear on this list,” Ely voters were told.

Local election officials traced the letters back to Panagopoulos before he could run the post-election ads, and persuaded him not to follow through on his vow to do so. But citizens who received the letters would have had no way of knowing that, and when he was able to look back at the voter file after November he saw, as he expected, that the threat of shaming was far more potent than the promise of praise.
Ely residents who received the postcard were nearly 7 points more likely to vote than those in the control group; turnout in Monticello and Holland increased by 4.7 points and 0.9 point, respectively. The impact in Monticello was only half as robust as the strongest of Grebner’s Michigan letters but still nearly six times better than the traditional-style GOTV mail that had wanly reminded prospective voters of their civic duty.

Panagopoulos started reading from the expanding portfolio of research being assembled by behaviorally minded economists who had found that expressions of gratitude helped to stimulate what they described as prosocial behavior. (In one field experiment,
two psychologists found that restaurant servers whom they directed to write “thank you” on their bills received larger tips from customers.) Panagopoulos, then teaching at
Fordham University, set up an experiment to take place in a New York City Council special election scheduled for February 2009. He identified single-voter households who had participated in the city’s last municipal election just over three years earlier, and sent around two thousand of them a postcard thanking them for having done so, and included a reminder about the upcoming special election. Those who received it ended up voting at a rate 2.4 percentage points higher than a control group receiving no contact. (Another group got just a postcard with an election reminder but no expression of gratitude; it had barely any impact on turnout.)

But most amazing to Panagopoulos was the silence. His surveillance hadn’t triggered any response—no disgruntled local election officials or righteous local television crews or death threats. When news of this trickled back to East Lansing, Grebner’s satisfaction at the influence of his approach was tempered by disappointment that decorum seemed to be winning the day. “We’ve now found forms that are nearly as effective that don’t turn people ballistic,” he says. “Although it still turns out that making them ballistic—boy, is that powerful!”

Malchow had had the idea of generating an “honor roll,” a roster of voters who never missed an election, and which would be sent to their neighbors. After Malchow learned about Panagopoulos’s work, the two partnered to compare their approaches in an experiment in New Jersey before voting there in 2009. Working with a labor-backed group defending Governor Jon Corzine against a challenge from Republican prosecutor Chris Christie, they sent out twenty-three thousand letters. Half contained Malchow’s honor roll, and half Panagopoulos’s declaration to recipients that “we hope to be able to thank you in the future for being the kind of citizen who makes our democracy work.” Both proved effective, with the “Thank You” letter increasing turnout by 2.5 points among recipients, and the Honor Roll by 2 points—the first costing just more than eleven dollars per additional vote. Since it was upbeat and congratulatory and threatened no future surveillance, there seemed to be no downside at all. When Malchow presented the New Jersey experiment at the Analyst Institute, he finally saw his
excitement reflected in his audience. “People lit up about that,” he says. “Because anyone can send a letter that says ‘thank you for voting.’ ”

Lighting up an Analyst Institute luncheon no longer amounted to impressing a small group of Malchow’s geek peers, but was now a method of directly inserting a new idea into the campaign plans of leading national Democrats. Obama’s inauguration had ushered in a new Democratic establishment in Washington, its rise bringing the data-driven crowd in from the outside. Hundreds of people were now cycling through the lunches and weekend retreats, an expanded audience that perhaps counterintuitively made participants even more eager to share their private research. Where better to show off? Many of the young staffers who had learned analytics on the Obama campaign had moved into top party jobs and reviewed the Analyst Institute presentations with commensurate authority and budgets.

When Malchow delivered the New Jersey results, it caught the attention of an operative from the Democratic National Committee, the same organization that Malchow had vainly tried to convince of the value of testing pre-election mail since Bill Clinton’s first presidential campaign. In the spring of 2010, before a special election for a Pennsylvania U.S. House seat, the Analyst Institute advised the DNC on its own version of the thank-you test. The experiment found that having the state’s popular senator Bob Casey tell voters on his letterhead that “our records indicate that you voted in the 2008 election”—and thank them for their “good citizenship”—helped nudge them to the polls this time. Analyst Institute members who had scoffed at harder-edged versions of social pressure were now eager to use it in the field. The approach may have been psychologically manipulative, but it no longer felt that way. “Volunteers were excited to deliver a thank-you message,” says Regina Schwartz, the institute’s outreach director.

By election day 2010, tens of millions of social-pressure mailers, in slightly different versions, were sent out by campaigns in both parties. Many had no ties to Gerber and Green or the Analyst Institute world, but merely took their core finding that social pressure worked and improvised.
In one case,
Utah Republican Mike Lee’s Senate campaign e-mailed supporters with a list of other voters in the same precinct who had a record of turning out in presidential elections but not in off years. “These voters do not understand the importance of mid-term elections and the direct impact their vote can have on our state,” the e-mail from the Lee campaign read. “We need to inform these voters!”

Suddenly, the use of social pressure as a turnout trick was so widespread that Gardner was describing it as “the hula hoop of American politics.” Even though the right did not have anything like the Analyst Institute to distribute such research in the form of simple recommendations, new techniques and tactics still moved from one operative to the next, or outward through the influence of officials at the party campaign committees. Whoever drafted the letter for Mike Lee’s campaign would probably pass it on to dozens of other Republicans by 2012, when they could use it to rouse their voters. Any competitive advantage the left had gained would likely fade, and as voters became aware of the letters the psychological impact of receiving one would weaken. The pressure was on Malchow to find something new.

SIX MONTHS LATER
, Malchow eased his new Chevy Volt out of the garage of his suburban Virginia home, whose backyard tumbled down onto the banks of the Potomac River, and drove to the Service Employees International Union headquarters near Dupont Circle in Washington. Malchow had purchased the electric car for all the usual lefty reasons, but it was not a sense of social responsibility that most tickled him about being a Volt driver. When the Chevy dealership offered Malchow a chance to participate in a Department of Energy program that would install an advanced recharging station in his home at no cost in exchange for being able to analyze the data it collected about his driving habits, he said yes without hesitation. Malchow has a guileless disinterest in privacy concerns when a
trove of new data hangs in the balance, and he was plainly pleased to have stumbled into an experiment, even as a subject.

These days, Malchow’s focus was on promoting a young-adult fantasy novel he had coauthored with his dyslexic teenaged son and beginning work on another. The challenge of selling a youth genre book fascinated Malchow, and he gave the impression of being more satisfied by drafting the marketing plan than he had been by the narrative itself, which grew out of the bedtime stories he began concocting with his son, Alex, when the boy was eight. Alex was now a high school football player, and at book readings he evinced a visible discomfort at being continually implicated in this vestigial preadolescent project, wincing as his father volubly declared his plans to pursue a sequel.

It was a Wednesday morning, and Malchow was in the unusual position of not having an office to go to. The previous November, on the eve of the midterm elections, he announced he was disbanding MSHC Partners, which had been one of Washington’s most successful consulting operations for two decades and a preferred mail vendor for four consecutive Democratic presidential nominees. (He had worked for Hillary Clinton in the 2008 primaries.) Malchow was accustomed to the rough-and-tumble of the business, but for him it had finally gone too far. He decided he would get out of the day-to-day of campaigns, disturbed that they often consisted of little more than strategizing how “to smear some poor guy with different beliefs.”

Many who received Malchow’s e-mailed announcement were shocked by it, but they rolled their eyes when they got to the part where he declared that “politics is not what it was when I started 25 years ago.” His claims of a conversion were disingenuous, some said, rumoring instead that he was easing out of the business for commercial reasons and grasping for a noble rationale to cloak his desperation. Even as Malchow insisted that 2010 had given the firm among its biggest revenues ever, there were signs that his business model had grown unsustainable. A major problem was the way Malchow’s instincts for entrepreneurship and self-promotion could frequently be at odds with each other. He would eagerly run experiments
and research projects to develop new techniques, often at significant cost, but then instead of guarding the findings for competitive advantage would rush to present them at Analyst Institute meetings. “The repercussions were that other people copied and sold it,” says Joel Rivlin, the director of analytics at Malchow’s firm. “Hal made his money. He always wants to push the envelope and bring everyone else with him because he thinks we can always do this stuff better.” Meanwhile, the firm’s corporate infrastructure grew so large (a five-person HR department, for example) that the roster of clients had to grow every year just to cover the overhead, a demand that employees described as “feeding the beast.”

Malchow noted in his announcement that he did intend to maintain one political client “I especially admire.” Everyone who knew the contours of Malchow’s enthusiasms understood he was referring to Women’s Voices Women Vote. The group’s limited focus on increasing participation and not persuasion meant it was always working on the easiest thing in politics to cleanly measure—the electoral rolls keep good track of who’s registered and who turns out—and so nearly from its outset Gardner decided that Women’s Voices would impose a sense of metric accountability on its operations. “We’re not big on exhortation,” she says. “We’re big on proving.” The group’s annual summits, a parade of academics and consultants showing off their latest research breakthrough on a series of PowerPoint slides, were part of the identity Gardner had worked to build as she fought to engage donors who had many other liberal organizations angling for their dollars. “Some of it is just Page showing off, for the community: here’s what we’re up to,” Malchow said. “Bring people in so they’ll do more stuff with you.”