Narrowcasting the 2012 Election

Why media polls that slice and dice the electorate could miss the big picture

NEW HAMPSHIRE — With Mitt Romney’s hold on the GOP nomination becoming too obvious to deny, horse race enthusiasts in the political media have quickly shifted to handicapping the general election. Unfortunately, their recent focus on key states and demographics—in particular, the effects of the contraception controversy on women voters in battleground states—threatens to obscure more fundamental factors that are most likely to shape the outcome of the campaign.

On Monday, USA Today and Gallup released a poll showing President Obama leading Romney 51%-42% in twelve swing states including New Hampshire. Last month, Romney was in a statistical tie with Obama, 48%-46%, in those same states.

Explaining the shift, USA Today’s Susan Page pointed in particular to Obama’s gains among “women under 50,” a important group of potential swing voters. She writes: “In mid-February, just under half of those voters supported Obama. Now more than six in 10 do while Romney’s support among them has dropped by 14 points, to 30%.” Page quotes Sara Taylor Fagen, a GOP strategist, attributing Romney’s drop to “[t]he focus on contraception” in media coverage and political debate.

The poll results soon attracted widespread attention, including among influential media outlets here in New Hampshire. The results were covered in online and print stories in the Boston Globe (here and here). The poll was also cited in a Reuters story by Alister Bull that ran in the New Hampshire Union Leader. Like Page, Bull cites statements by Obama campaign officials attributing the drop to the contraception debate among GOP candidates.

Yesterday, the national press dove even deeper into the details of the poll. Karen Tumulty and David Nakamura of the Washington Post focus on Obama’s standing among independent women who were surveyed, while Ron Brownstein of National Journal differentiates Obama’s support among women based on whether they have a college degree and their racial background. The contraception debate was raised as a contributing factor in both stories. According to Tumulty and Nakamura, Romney pollster Neil Newhouse “said the candidate’s problems with women probably represent collateral damage from the arguments that women have been hearing about contraception and other social issues.”

How much does any of this matter in electoral terms? It’s certainly noteworthy if Romney’s support has declined significantly among women, but the election will likely be won or lost based on the national-level vote swing, not shifts in President Obama’s support among women under 50 in swing states more than seven months before the election. Journalists are looking in the wrong place.

First, while the decision to run a poll exclusively in swing states is a great marketing tactic, it’s not clear how much we can learn from such a poll so early in the campaign. Even national trial heat polls are only somewhat accurate at this point—there is no reason to think that polling in swing states will be any more predictive. Indeed, state-level vote swings between elections have become more similar over time, suggesting that national-level polls may be more likely to predict which states become competitive than the converse. (For instance, as Barack Obama performed well in 2008, he put states like Indiana into play that were thought to be out of Democrats’ reach.)

A similar argument applies to the role of women and the effects of the contraception debate. They do play a key role in the Democratic coalition, but the media’s focus on specific demographic groups tends to obscure the role of overall vote swings between presidential elections. For instance, both presidential campaigns fought fiercely for women’s votes in 2008. Much was made of Sarah Palin’s potential appeal to women when she was added to the GOP ticket, but initial post-convention gains in John McCain’s standing among women quickly dissipated in an unfavorable environment for Republicans. In the end, exit polls found that Barack Obama improved by five percentage points on John Kerry’s vote total overall (53% to 48%) and among both men (49% to 44%) and women (56% to 51%). As a result, the gender gap—the difference in support for the winning candidate between men and women—remained seven percentage points, the same as in 2004 (PDF).

In reality, presidential election outcomes can almost never be attributed to a shift in a single demographic group. Likewise, most campaigns are decided by the popular vote, not the details of the Electoral College. For both reasons, journalists should keep their eye on the big picture. While forecasting models are hardly perfect, they have persuasively shown that presidential elections are shaped by fundamental factors like incumbency and the economy, which tend to move demographic groups roughly in parallel. Obama appears to be overperforming among women now, but campaigns tend to move voters toward the outcomes we’d expect given the fundamentals. The implication is that Romney’s standing among women is likely to recover somewhat. As I recently noted, campaign shocks to candidates’ standing in general election trial heats are largely transitory at this stage of the campaign (link requires subscription). Though the gender gap will persist, Republican women and GOP-leaning independents are likely to find reasons to return home after contraception leaves the news, Romney’s rivals stop attacking him, and the conventions remind them of their partisan loyalties.

In the end, the contraception controversy is a real policy issue with vastly more significant consequences than the Etch-a-Sketch gaffe. But like that episode, its impact on the presidential campaign will likely soon be forgotten.

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.