Victory Lab

Stuff Some Envelopes, Then Ask Questions

Reporters should have to volunteer for a campaign before covering one.

Obama makes a phone call to a supporter during a visit to the Obama for American campaign field office
Obama makes a phone call to a supporter during a visit to the Obama for American campaign field office in Port St. Lucie, Fla., in September during the second day of a two-day bus tour across Florida

Photograph by Saul Loeb/AFP/Getty Images.

I’ve been out this past week talking about my new book, The Victory Lab. In interviews, the conversation often leads to an essay I wrote for the New York Times a few Sundays ago, in which I argued that campaigns had become so technically and intellectually sophisticated that journalists were no longer capable of competently covering the horse race. A number of interlocutors have asked the obvious question: What should journalists do to adjust to this new reality?

I’ve struggled a bit when faced with that question, and have responded with the admittedly fuzzy suggestion that the media assume more humility when describing campaign dynamics. This amounts to repeating a point I made in the closing paragraph of the Times essay: “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.”

I’ve felt a little unsatisfied with this advice and suspect that those asking the questions have, too. So I thought I would depart from amateur epistemology and offer one highly practical — if for now implausible —short-term recommendation to improve the quality of horse-race coverage. Journalists should work on campaigns.

To be clear: I don’t think our field requires more Stephanopolouses or Kristols or Reid Cherlins, who have been spokesmen, or speechwriters, or message strategists at a relatively high level. We need working reporters who have spent time inside a field office and have the comfort with the street-level politics that an engaged activist would develop after a few months of regular volunteer shifts on a modern campaign.

Journalists regularly write things that betray their distance from the daily workings of politics, like this paragraph that appeared on Page One of the Sunday New York Times a little more than a week ago, in a big-think piece assessing the post-convention state of play:

Presidential races take place on many levels, some easily visible, others more shrouded. As the clock runs down, both sides make tough decisions about which states to compete in and which to abandon. Advertising themes get tested and changed as strategists hunt frantically for the right appeals, and get-out-the-vote teams target wavering voters with tailored messages.

That paragraph is so vague that it’s shocking there could be anything inaccurate in it, but its last sentence appears to betray a fundamental misunderstanding of what “get-out-the-vote teams” do. Their work, known as GOTV, focuses on mobilizing existing supporters who are not regular voters; campaigns assume that these people need reminders, nudges, information on polling-place locations, and offers of rides to the polls. Such targets are not by any measure “wavering”; campaigns isolate these citizens for GOTV contact because they believe them to be a likely vote if properly mobilized. There is also very little interest in developing “tailored messages” for them, because campaigns are not aiming to change their opinions as much as alter their behaviors.

A journalist who had witnessed the process by which names get selected for GOTV walk sheets or call lists would be unlikely to fall victim to such confusion. Every name included in a list of GOTV targets would belong to someone who had told a campaign at some point that she was a supporter, but whose vote history indicated she could not be counted upon to turn out of her own accord. Nowadays, campaigns use statistical modeling to classify GOTV targets based on available voter information, which allows them to find targets without having to question each person directly.

The media tends to cover the mechanics of field contact only in the final weekend before an election. At that point, covering door knocks is rarely treated as a task for a senior journalist, who is invariably working on a larger strategic story assessing the pre-election state of play. Instead, GOTV is treated as the political world’s version of the solstice, a recurring and predictable event worth noting but about which there is little new to be said. The Saturday before the election, television cameras head to campaign offices to film field staff handing out clipboards to canvassers, or to phone banks to watch volunteers place calls, so producers have b-roll to accompany the inevitable voice-over: Campaigns rushed frantically to reach their voters in the closing hours of the campaign …

But turnout is often the first thing that analysts within a campaign consider when launching an election bid. Sometimes before a dollar has been raised or the lease signed on a headquarters, those analysts are assessing vote goals: What will it take to win, and where will the votes come from? How many are guaranteed through reliable base voters? How many decided supporters are there who require an extra push to the polls through GOTV operations? Only if the sum of those two categories falls short of the “win number” does a campaign start looking for the other voters—either undecideds or soft supporters of the opposition—it will need to persuade to vote for their guy. Without appreciating how a campaign arrives at its assumptions about whom to mobilize for turnout, it is difficult to make much sense of its persuasion strategy.

What if journalists actually developed a working knowledge of those mechanics and the tools campaigns used to engineer them? It doesn’t take much to dramatically increase your base of knowledge about voter-contact tactics, which often reveal more about a campaign’s thinking about where its votes will come from than the latest Web ad or polling memo released by a communications department.  A lot of media have written about the Obama campaign’s new mobile canvassing app, but few have asked a central question about its underlying purpose: Why is the Obama campaign using it to send existing volunteers to recruit other volunteers instead of hunting for new supporters?

So a modest proposal: newsrooms develop a version of a study-abroad program, placing their reporters in campaign field offices for a month during the summer of an election season. It’s time that they see the place where campaigns interact with real people, by asking the questions on phone-bank scripts, entering the answers into databases, then seeing how that information shapes decisions about which voters to call or visit next. (Full disclosure: I volunteered for Democratic candidates in New York starting when I was 12, but have not worked in campaigns since I first got involved in journalism.) My guess is that journalists who spent even a few weeks in this world would pose wildly different questions the next time they sat down with Jim Messina or Stuart Stevens.

The goal would not be to gather intelligence about a particular candidate and his tactics but to build institutional knowledge that could help to re-center journalistic understandings of what a campaign actually does. To assuage concerns about bias and conflicts of interest, newsrooms could assign reporters to work in races away from the ones they cover: the Richmond Times-Dispatch correspondent gets detailed to an Oregon mayoral campaign, the Nevada radio reporter to a Maine ballot initiative. Assignment editors at the Washington Post and Politico and NBC News would randomly dispatch their reporters so they’re split evenly, half campaigning for Democratic candidates and half for Republicans. While it would be great if they could slot into presidential campaigns, it’s by no means necessary. Many of the essential tools used by campaigns for organizing and marshaling voter data have become so universal that a national political journalist would learn plenty from being exposed to a competent modern campaign for state legislature or county judge.

The media have become so fixated on neutrality that we have become detached not only from the ideologies and philosophies of the people we cover, but their methods, too. Many have noted that the turn toward rigorously empirical campaigning looks a lot like the one that has changed baseball over the last decade, as described in Michael Lewis’ Moneyball. But political reporters are in even worse shape than the sports press was. Baseball beat writers may have had trouble appreciating and assessing the quantitative revolution taking place in front offices because they lacked the requisite statistical expertise. But political reporters are actually far more poorly positioned to document similar transformation in our fields. Nearly everyone who writes about baseball has held a bat and a glove at some point, and some—no matter how old or physically unfit they may be now—still play catch, dabble on a softball team, or coach Little League. The first step toward appreciating why a general manager might prize a statistic like a pitcher’s WHIP is understanding what it takes to throw a baseball 60 feet, or the choices a batter faces at the other end.

Journalists have rarely developed a comparable familiarity with the fundamentals of the game we cover. We haven’t held a canvasser’s clipboard and seen the empty fields that are to be filled in after a conversation with a voter at his doorstep. How could we possibly understand what would go into a statistical model predicting her vote intention, or a randomized field experiment to test her persuadability? It’s time for political beat reporters to pick up a bat and see what it’s like to take a swing.