Why Campaign Reporters Are Behind the Curve

Slide Show

It becomes popular around this time of year to lament the fact that media coverage treats the presidential campaign as little more than a “horse race.” Journalists, this line of argument goes, choose to fixate on which candidate is a superior campaigner or savvier strategist, not on who has sounder ideas or is better prepared to govern. From time to time, the journalists themselves concede that to maintain daily or hourly tension in the contests they promote, they have little choice but to elevate minor poll shifts into major developments.

But the reality about horse-race journalism is far more embarrassing to the press and ought to be just as disappointing to the readers who consume our reporting. The truth is that we aren’t even that good at covering the horse race. If the 2012 campaign has been any indication, journalists remain unable to keep up with the machinations of modern campaigns, and things are likely only to get worse.

“My view is that there’s nothing that’s secret in campaigns anymore — but that doesn’t mean everything is understandable in a campaign,” says Terry Nelson, who served as John McCain’s campaign manager in 2008. “The ability of campaigns to run circles around journalists in some places is strong, and it’s not healthy.”

I covered the 2008 election for The Boston Globe, filing articles that I hoped would rise above the superficial and ephemeral poll-driven reporting that I had been trained to disdain. But after spending the last two years reporting on the scientific revolution that is quietly reshaping politics, I realized how much of the story my colleagues and I had missed.

Over the last decade, almost entirely out of view, campaigns have modernized their techniques in such a way that nearly every member of the political press now lacks the specialized expertise to interpret what’s going on. Campaign professionals have developed a new conceptual framework for understanding what moves votes. It’s as if restaurant critics remained oblivious to a generation’s worth of new chefs’ tools and techniques and persisted in describing every dish that came out of the kitchen as either “grilled” or “broiled.”

“When I went to work for my first campaign, in 1994, I was actually surprised at how journalists tended to think one step ahead where campaigns are four steps ahead,” says Joel Benenson, a former newspaper reporter who now serves as President Obama’s chief pollster. “Think of it as a level-five player in chess and a level-eight player in chess. You had people covering campaigns who are at the mercy of the grandmasters of politics.”

The gap between journalists’ desire to cover the political game and our ability to do so has only widened since Mr. Benenson changed careers. Campaigns have borrowed techniques from the social sciences, including behavioral psychology and statistical modeling. They have access to private collections of data and from their analysis of it have been able to reach empirical, if tentative, conclusions about what works and what doesn’t.

But we have done little to update our thinking. Until about 2000, we were able to keep pace with major innovations in the political world. When new tools were developed for measuring public opinion — whether it was tracking polls, focus groups or the so-called dial sessions that measured a voter’s instantaneous response to a video — news organizations could replicate them. When taking stock of a race’s dynamics, journalists reviewed many of the same types of research that sat on a campaign manager’s desk.

We can’t keep up with the revolution in political information gathering.

In the years that followed, campaign analysts began to pull in reams of new data on individual voters. Politicians have always looked for ways to communicate differently with niche audiences on issues of narrow concern, but they had been stuck approaching them in terms of geographic zones or familiar demographic subgroups. Now campaigns had access to all sorts of new demographic and lifestyle markers, like lists of people who purchased religious material or had gun licenses or had recently taken a cruise.

Breathless, and often fact-free, stories about “data mining” and “microtargeting” soon became plentiful. But few journalists had access to any of the campaigns’ data, or even much understanding of the statistical techniques they used. We found ourselves at the mercy of self-promoting consultants who described how they were changing politics by ignoring stodgy old demographics and instead pinpointing voters according to their lifestyles. We played along, guilelessly imputing new mythic powers to microtargeting. In many retellings, data analysis became the reason George W. Bush was re-elected.

Microtargeting was at once less directly influential, and more fundamentally disruptive, than these analyses suggested. The most colorful commercial variables that appeared prominently in journalistic accounts of microtargeting — whether someone drank gin or drove a Subaru — were never of much value on their own. It was the combination of hundreds, sometimes thousands, of data points that offered value: algorithms could weigh previously imperceptible relationships among variables to predict political attitudes and behavior.

Now, instead of defining voters by a handful of self-evident attributes like rural Hispanic Democratic men or non-college-educated white seniors, campaigns could group individual citizens according to segments or scores that reflected gradations of predicted habits — of how likely they were to turn out to vote or to support a specific candidate. They could be aggregated into what campaigns call a universe — targets for the same persuasive media or get-out-the-vote drive — not by visible demographic commonalities but because they were projected to behave in similar ways.

Five days after Mitt Romney selected Paul D. Ryan as his running mate, Mr. Obama’s campaign released a public memo by Mr. Benenson with the title “Romney’s Choice of Ryan Falls Flat.” In it, Mr. Benenson reviewed survey data from pollsters like Gallup and Rasmussen to argue that Mr. Romney had not realized the same post-running-mate bounce as previous nominees had. The media duly covered the memo as news, under headlines like “Obama Pollster Says (With Data) Ryan Pick a Dud.”

In his Aug. 16 memo, Mr. Benenson wrote that “Ryan has had virtually no impact on Romney’s position.” That may have been a fair conclusion to draw from the public horse-race polls available then. But the publicly available data Mr. Benenson cited relied on an entirely different sampling methodology than the ones based on microtargeting scores that Mr. Obama’s polling operation actually uses to guide campaign strategy. Because their proprietary data was more varied and nuanced than Gallup’s, Mr. Obama’s advisers also knew then that it was too soon to assess Mr. Ryan’s impact. His selection could still cause dramatic changes to the contours of the contest without an obvious disruption to Mr. Romney’s standing in the horse race.

Indeed, the telling numbers wouldn’t be polls but the individual probability scores that Mr. Obama’s targeters developed (and update weekly) to predict how likely each voter in the country is to support him. As the scores adjusted to reflect post-selection opinion, there was the prospect that they could show a tranche of Romney backers (likely older whites) incrementally weakening in their support for the Republican ticket. Obama tacticians would relish the news: it would signal the emergence of a new persuasion universe where the president could play offense and force Mr. Romney to defend against defections.

Contrary to what Mr. Benenson’s public memo suggested, the Obama campaign wasn’t merely concerned with those who had already moved because of the Ryan pick. Chicago was already one step ahead, tracking those who may have just become susceptible to future movement. Once the campaign had identified those voters, it could start communicating with them, either through individually targeted contact like mail, phone calls and Web ads or niche media, which often elude the attention of the national political press.

“All journalists have one channel and all campaigns have one hundred, between Internet, TV, e-mail,” says Samuel Popkin, a political scientist at the University of California, San Diego, who has been a strategist for four Democratic presidential candidates. “They’re out there thinking, ‘What should we put in the goldbug newsletter or the Hadassah weekly?’ and the reporters are all thinking about what you’re putting in their paper.”

Journalists tend to mistake the part of the campaign that is exposed to their view — the candidate’s travel and speeches, television ads, public pronouncements of spokesmen and surrogates — for the entirety of the enterprise. They treat elections almost exclusively as an epic strategic battle to win hearts and minds whose primary tools are image-making and storytelling.

But particularly in a polarized race like this one, where fewer than one-tenth of voters are moving between candidates, the most advanced thinking inside a campaign is just as likely to focus on fine-tuning statistical models to refine vote counts and improve techniques for efficiently identifying and mobilizing existing supporters.

“There’s a lot that goes on in a campaign that reporters never really get at,” says Mr. Benenson. “There are a lot more things at play.”

Failing to appreciate those nitty-gritty tactics can mean missing the bigger strategic story altogether. The most examined inside-baseball campaign topic of 2011 concerned whether Mr. Romney would cede the Iowa caucuses to his rivals. If he threw himself into Iowa and lost, the thinking went, he would reveal weaknesses in a front-runner’s candidacy, like limitations that his faith or record placed on his reach among conservatives.

For nearly the entire year, all signs pointed to the notion that Mr. Romney was holding back for fear of experiencing a repeat of his embarrassing loss four years earlier. The candidate rarely visited the state, lacked a Des Moines headquarters, skipped the Ames straw poll and did not air a single advertisement.

Within Mr. Romney’s campaign, however, his options were never seen as the binary choice presented by the “Will Mitt make a play for Iowa?” media parlor game. While journalists waited for physical manifestations of a Romney “ground game” to materialize, Mr. Romney deployed statistical models to track Iowa supporters and current vote counts for his rivals. It amounted to a largely invisible 21st-century upgrade to the traditional infrastructure of offices, phone banks and staff that most journalists visualized when they tossed around the term “organization.” Only six weeks before the caucus did Mr. Romney unveil the trappings of a traditional caucus campaign.

On election night, when Mr. Romney was declared the caucus winner, the press treated it as validation of his ability to compete on turf dominated by party activists. But Mr. Romney had not significantly expanded his support in four years. While maintaining a fiction to guard him in case of an Iowa loss, his aides had been diligently counting votes until they had the confidence to know that external dynamics had transformed a losing coalition in 2008 into a winning one. (Mr. Romney’s victory was actually later reversed into a 34-vote loss to Rick Santorum.)

Mr. Romney had exploited the inefficiency at the core of contemporary campaign coverage: the press’s fascination with strategic calculations and gamesmanship well exceeds its ability to decode the tactics underneath. We may be covering the horse race with more bytes and airtime than ever before, but we’re looking at the wrong part of the track and don’t know how many legs are on a thoroughbred.

This failure to properly cover the contest should disappoint more than those who want to follow the presidential race as fans, relishing it as competition. The campaign horse race may be our great quadrennial national sport, but how candidates win matters. The coalitions they build, their reliance on party structures and activist networks, the places they choose to spend money and how they allocate their time and other resources all help illuminate the decisions that they will make after taking office. To understand how they will govern, we need to understand how they run.

The smartest people I talk to in political campaigns — the ones who spend the most time in the company of advanced data and sophisticated experimentation — are also the quickest to concede how little we ultimately know about what it takes to win. For them, empiricism breeds uncertainty. Only by knowing what is measurable can we appreciate how much isn’t, and be honest with readers about the fact that everything else may have to remain a mystery.

Sasha Issenberg is a columnist for Slate and the author of the forthcoming book “The Victory Lab: The Secret Science of Winning Campaigns.”