The American Connection
François Hollande has embraced the tactics of Obama’s 2008 campaign—and it might win him the French election.
The three decided they wanted to send some volunteers to those places. They were familiar with a decade’s worth of field experiments, pioneered in the Yale lab of political scientists Alan Gerber and Don Green, that had measured the relative effectiveness of specific get-out-the-vote tactics by measuring their impacts on voter turnout. (Green is now at Columbia.) In the fall of 2009, Liegey, Muller, and Pons approached the Socialist leader in the capital region, Maxime des Gayets, with a request to use the following spring’s regional elections as a trial. Des Gayets was initially skeptical, describing himself to Mediapart as “rather suspicious of the Obamamania in the air,” but eventually gave them 80 party activists to deploy as canvassers.
They identified eight zones, seven suburban cities and one Paris arrondisement, that had high rates of nonvoting but had supplied large margins for left candidates in recent elections. Because French electoral rolls make it difficult to segment out individual households—apartment residents, including the large public-housing complexes that stud the Paris suburbs, are usually all listed at the same address—the three knew they would have to assign treatments not at the individual level, but by building. They randomly assigned 675 buildings to get a visit from teams of party canvassers in the weeks before the election, and 675 to go unvisited as a control group. (The two groups covered around 24,000 registered voters.)
After voting, local French authorities open the electoral rolls to public scrutiny for 10 days, and Liegey, Muller, and Pons found that their mobilization effort had had its intended effect. The control group turned out at a rate of 31.8 percent, while those whose doors had been knocked on did so at a rate of 35.9 percent. But the effects had been unevenly distributed. Using the few pieces of information the rolls carried on each voter, they found that place of birth was the crucial attribute in understanding why some voters had responded more positively to the get-out-the-vote visit than others. Among those born in mainland France there had been little movement, but among those born abroad or in the country’s overseas territories the effect was substantial. It is a staple of French election-season chatter to discuss the ways in which the country’s immigrant communities, especially those from North Africa, are politically marginalized, but it turned out in some cases that all it took was a knock on the door to turn “abstentionists” into voters. “No one ever told them that voting was important, that their vote mattered,” says Liegey.
Liegey and his colleagues pushed party officials to make door-to-door mobilization a focus of their 2012 campaign plans. “To win an election, it is often more efficient to mobilize voters from one’s own camp at risk of abstaining than it would be to try to persuade the undecided or voters on the other side into voting for you,” they wrote in an report for the Socialist Party’s internal think tank. The findings found an enthusiastic backer in Vincent Feltesse, a digital strategist who had knocked on doors as a local candidate and saw promise in expanding the practice to the party volunteers who usually spent their time indiscriminately stuffing mailboxes with party platforms or handing out leaflets at transit stops. The Obama connection was particularly appealing to Feltesse: He would enlist Blue State Digital, the firm responsible for Obama’s Web operation, to engineer Hollande’s online presence.
Lessons from the Obama experience had already given the Socialists one structural advantage as they prepared to face Sarkozy in 2012. Inspired by the lively 2008 Democratic primary season, the party became the first in France to choose its nominee not through a ballot open only to party insiders but through one that was open to any voter who wanted to participate at a local polling place. The October vote proved an organizing boon to the party: Of the 3 million voters who turned out, about 700,000 offered to sign in with their names and contact information—more than five times the number of activists then in party files. After the New Year, Hollande’s campaign used robocalls to encourage these primary voters to volunteer.
On the heels of this success, the troika were soon installed in a corner of Hollande’s headquarters as the door-to-door team, and rather arbitrarily set the grand goal of visiting 5 million households. They set out to divide the French landscape for maximum impact, calculating an index of mobilization potential for Hollande that multiplied a precinct’s abstention rate by its combined vote for left-of-center candidates in elections since 1998. They then ranked the country’s more than 60,000 polling stations according to the index. Each polling station covered about 1,000 voters, so they took the top 5,000 areas and made them target precincts for mobilization. (The designation covers many largely minority areas in the suburbs of cities, but also pockets of posh left-leaning Paris neighborhoods where the index pointed to a potential for squeezing out extra votes.) The troika began assigning 70,000 volunteers to these promising patches, and set up a reporting hierarchy to monitor their progress. By Friday, 3.4 million doors had been knocked on, and around half of them opened, as the reports describe the act of actually interacting with a voter.
When a door opens, volunteers are careful about what they say. Interrogating a stranger about how he or she plans to vote, as Muller did on Obama’s behalf on New Hampshire doorsteps, would be off-putting in France, he believes. Instead Hollande’s canvassers are instructed to coyly ask what their targets think of the candidates as a way of opening up a conversation.
There are limits to how much room there is for Obamamania in the air of French politics. The country’s rigorous privacy rules restrict the ability to collect the types of personal information necessary to sort voters and track their individual turnout, and the political system lacks the money and statistical expertise to match what is available to American targeters. (A former pollster to Sarkozy, inspired by an American trip in which he met Karl Rove, spent a year trying to marshal resources to assemble a rudimentary national voter file but eventually abandoned the project after failing to recruit financial backers. He is hopeful that the party can take on the project in time for the next presidential elections, in 2017.)
No amount of commitment to the principles of mobilization will ever be able to fully translate “get out the vote” to French. Election laws require campaigns to wind down all activity at midnight on Friday, which means that for the last two days there is not the typical American frenzy of final pre-election reminders or the dispatching of vans and buses to drive voters to the polls. “It’s telling that the French perception of voting,” says Pons, “is that you get all your information and then you get 48 hours to consider it and make up your mind.” The troika plan to spend their weekend playing sports and catching up on rest, before restarting Monday on a two-week sprint to the second round and their quest to reach 5 million doors.
Sasha Issenberg is the author of The Victory Lab about the new science of political campaigns.