The over-covered image war

Journalists are exaggerating the risk that Mitt Romney will be "defined" early

The message war in the presidential election got underway in earnest last week, with the Obama campaign releasing a new attack ad and super PACs on both sides announcing their own big buys. If you believe some prominent voices in the political press, the stakes in these exchanges are already extremely high—especially for Mitt Romney, the likely GOP standard-bearer.

Many reporters and commentators have suggested that Obama could “define Romney” in the minds of the public—or at least, that he’s trying to do so—and thus establish an image now that would prevent Romney from winning in November. Evidence for the idea that we’re now in a crucial part of the campaign, however, is generally lacking.

Perhaps the most extensive version of the claim was advanced by Richard W. Stevenson of The New York Times, who wrote last week that “the months between the end of the primary season and…the conventions are an especially perilous period for candidates in Mr. Romney’s position,” since they are then “most susceptible to being defined on terms other than their own.”

Stevenson’s framing reflects the thinking among political operatives who were interviewed by the Times’s Jeremy W. Peters, who described “a belief among political strategists on both sides that they have a small window of time to frame the election” before summer reruns and the Olympics. For instance, Democratic strategist Carter Eskew told Peters that, “The first ads that are run are in many ways the most important because the mind is the most open and uncluttered at that point.”

In reality, however, most Americans’ minds aren’t “open and uncluttered” about Romney; nearly all have already formed an opinion about the presumptive Republican nominee. The latest Gallup poll finds that 91% of Americans already have an opinion of him. And despite the dangers to his image that Stevenson cites, Gallup found that Romney’s favorable numbers are predictably improving as Republicans and GOP-leaning independents return to the fold (though these improvements have not yet shown up in other polls).

As a result of these pre-existing views and the passage of time, most of the messages that voters are receiving now about Romney will wash out over time. Research by the political scientists Christopher Wlezien and Robert Erikson on trial heat polls (gated) finds that pre-convention “news about the campaign affects voters but is eventually forgotten and thus has little impact on the final outcome.” (Peters’s story, to its credit, acknowledged up high “concerns that [early advertising] money will be wasted.”) And even if those messages do have a lasting effect on perceptions of Romney, the political scientist Larry Bartels has found that presidential “candidates’ images are largely epiphenomenal and have only a modest impact on election outcomes” (gated)—i.e., their images are more a consequence than a cause of how the candidates are faring.

It’s true, of course, that you can select unsuccessful candidates after the fact and claim that their image ratings were the reason that they lost. Stevenson cites Michael Dukakis, Bob Dole, and John Kerry as out-party candidates who “failed to establish strong positive images during this period and allowed their opponents to brand them in ways they never overcame.” However, these are all retrospective judgments. At this point in the 1988 campaign, the media focus was on George H.W. Bush’s perceived struggles to define himself. Moreover, all three candidates that Stevenson cites lost in economic circumstances that would lead us to expect the incumbent party to hold the White House. The state of the economy—and the extent to which Obama is blamed for it—is still most likely to determine Romney’s fate. As Roll Call’s Stuart Rothenberg points out, “when November rolls around, the question of who Romney is might not be nearly as important to voters as how well Obama has done.” In short, it matters far more how Obama is defined than Romney, whose image is likely to matter only on the margin.

Of course, it’s possible for anything to change the outcome of a campaign if the vote is close enough, but the media should treat the battle to “define” Romney with far more skepticism. More generally, reporters should refrain from overstating the importance they place on early-stage campaign squabbles. According to Wlezien and Erikson, the real action comes in the final 100 days, which is when campaign shocks start to “persist to affect the outcome of Election Day” (typically in the direction we would expect given the state of the economy).

In other words, the conventions aren’t just the beginning of the fall campaign; they mark the beginning of the period when campaign events really start to matter. In the meantime, why not help voters learn about the campaign promises that the candidates are making and the agendas they are proposing as they roll out their general election platforms? The horse race won’t be decided now, but the agenda of the next president will be.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Brendan Nyhan is an assistant professor of government at Dartmouth College. He blogs at and tweets @BrendanNyhan.