Conventional analyses of the 2012 presidential race paint a consistent picture: President Barack Obama’s data-driven, technologically savvy campaign dominated the Mitt Romney campaign, out-organizing on the ground and out-messaging on the airwaves. Yet an emerging body of research is challenging this currently accepted narrative, asking whether the election had an inevitable conclusion despite the endless resources and hours spent by both campaign teams.
The Early Advertising Game
Political scientists are in disagreement with Obama campaign advisors over the efficacy of early advertising, a strategy
in which Obama For America heavily invested. These ads drew the attention of Professor John Sides at George Washington University and Professor Lynn Vavreck of UCLA, who studied this strategy of early advertising in their book, The Gamble. Sides noted that “ads on the day before the election appeared to produce small but measurable gains in vote share,” yet “ads aired before that Monday—even the calendar week before the week of the election—had a much smaller effect.” He and Vavreck found back-loading advertising to be much more influential upon voters on Election Day.
That finding conflicts with senior Obama advisor David Axelrod’s account of the campaign. Axelrod explained to the HPR that early TV ads are effective because of both their timing and the more malleable audience. Axelrod also spoke about the relationship between the Obama campaign’s advertising strategy and national media coverage of the election. As the cycle progressed, the argument goes, and the press pays more attention
to the candidates, the influence of advertising begins to wane. In particular, “news overwhelms advertising after the nominating conventions,” according to Axelrod.
Indeed, political scientists have studied the optimal time to advertise during a campaign for several years. In one remarkable study, Texas Governor Rick Perry’s 2006 reelection campaign allowed the launch date and volume of television advertising to be randomly assigned in 18 of the state’s 20 media markets for a period of time. Professors from Yale, the University of Maryland, and the University of Texas, analyzing the data, concluded that “television campaign advertisements initially have a large and statistically significant effect on voter preferences,” but “just a week or two later, the advertisement’s effects have all but disappeared.” These results may be specific to the types of positive ads played during the Perry campaign, however.
There is agreement about the impact of sustained advertising campaigns, which characterized the 2012 presidential election. In an interview with the HPR, The Washington Post’s Reid Wilson wrote, “The Obama campaign took a major risk by spending money earlier than other campaigns in order to start telling the story of Mitt Romney, Bain Capital, and Massachusetts.” The story of an elitist, wealthy businessman who could not relate to ordinary Americans was a specific narrative “that ran through all of Obama’s advertising.” According to Wilson, the message was both believable and constant, which is why it proved so effective.
Vavreck is more skeptical, due to fact that this was not exactly a novel way to criticize Romney. “Other Republicans had labeled Romney as a rich, out of touch, plutocrat long before Obama did this in July, so I doubt those ads redefined Romney in a ‘new’ way,” he said. But he also made clear that sustained advertising can mean a sustained advantage, and Organizing for America had exactly the resources for this sort of campaign.
Bowdoin professor Michael Franz expressed that idea memorably. Franz, who believes Sides and Vavreck “undersell the potential effects of early ads,” argues that ads are “like drugs, in
a way. You’re trying to get the electorate high on you,” and then “they try and keep you high for the course of the campaign.” It comes as no surprise, then, that Axelrod made the case for the continuing effect of the early ads. “We defined the race and Governor Romney before the conventions, and he was digging out of that hole for the remaining two months,” Axelrod explained.
Technology and Voter Mobilization
Apart from their advertising strategy, the Obama team was widely lauded for their on-the-ground voter mobilization efforts, a strategy that also enjoyed success in 2008. Howell Raines, the former New York Times executive editor, articulated a common perception of the Obama team when he wrote in The Washington Post: “But the overriding lesson for future candidates lies in Obama’s computer-driven voter-mobilization machine.” He described the operation as a combination of “digital weaponry and an army of foot soldiers.” But how impactful was all that digital weaponry and that army?
Ryan Enos of Harvard and Anthony Fowler of the University of Chicago measured voter mobilization by comparing activity among voters who live in the same media market but are divided between battleground and non-battleground states. Although Enos emphasized to the HPR that their results are part of a continuing project and subject to minor change, their preliminary findings show that “there are campaign effects of 15.4 percentage points for registered Democrats and 13.8 percentage points for Republicans.” These numbers suggest that both the Obama and Romney campaigns ran effective on-the-ground strategies. “Despite all the celebration of the Obama campaign’s technological and other superiority, their campaign only had a 1.6 percentage point advantage over Romney in turning out party registrants,” he argued.
Of course, Axelrod pushed back, noting that even that 1.6 percent figure makes a large difference in battleground states. On the margin, he argued, these are the voters that win Ohio and Florida and thus the Electoral College. Yet even Axelrod compared the Obama campaign’s technology strategy to a “field goal team”—helpful, but not able to win the game singlehandedly.
How Much Credit Does the Obama Campaign Deserve?
Enos sees a psychological effect behind the effusive praise for the Obama campaign. He thinks that the Obama campaign probably “receives too much credit” due to what’s known as attribution error. In essence, people are likely to “attribute causes to things we see immediately in front of us,” regardless of their role in the actual outcome. To Enos, and many others, the outcome of the election was fairly predictable, but analysts love to point to specific strategies and tactics used by the campaigns to explain the results.
Drew Linzer was one person who had such highly accurate, Nate Silver-style election forecasting models long before the election. Linzer’s models combined historical prediction criteria (changes in GDP, for instance) with up-to-date polling data, and he correctly predicted all 50 states as early as June. But even that predictability and stability hasn’t convinced Linzer that campaigns are ineffective—he explained to the HPR that he believes campaigns can still matter.
Where does this leave us? The academic literature suggests in places that presidential campaigns aren’t as influential as they might seem in the news. Daily stories about the latest polling numbers due to a gaffe or specific advertisement certainly make for attractive entertainment, but also might be much ado about very little. Regarding Nate Silver, New York Times public editor Margaret Sullivan wrote, “A number of traditional and well-respected Times journalists disliked his work.” Sullivan also drew an interesting comparison, “Much like the Brad Pitt character
in the movie Moneyball disrupted the old model of how to scout baseball players, Nate disrupted the traditional model of how
to cover politics.” Sides himself has invoked Moneyball on the Times website, so perhaps journalism is ripe for its own transformation, just as campaigns have seemingly changed in the past few years.
We don’t have a final tally on the efficacy of presidential campaigns—we might never—yet parsing through the post-November accounts of Axelrod, professors, and the campaign press corps, we can get a better grasp on that elusive and important tally. The impact of presidential campaigns, like so many other things, is likely a story best told with moderation in mind.