« October 2007 | Main | December 2007 »

Our Latest From Iowa

November 19, 2007 5:50 PM

(Updated)

We’ve now posted both analyses of our new ABC/Post poll in Iowa, reporting results among likely Republican and Democratic caucus-goers alike, with fascinating results on both sides.

We report the surprising surge for Mike Huckabee in the state, remarkable for its intensity as well as its breadth. A Baptist minister, he's now leading longtime front-runner Mitt Romney by 2-1 among evangelicals, who account for nearly four in 10 likely participants in the Republican caucuses.

Among Democrats, the overall race has hardly changed, albeit with just enough movement to give Barack Obama a statistically significant lead over John Edwards – surely not welcome news for Edwards, who's run what’s been described as a make-or-break effort in Iowa.

The Obama-Hillary Clinton race remains quite close, among likely caucus-goers overall and also among the subset who say they’re "absolutely certain" to attend a Democratic caucus. But there’s more going on just under the surface – a shift toward Obama on his “new direction” theme, some significant blowback for Clinton on the issue of saying what she really thinks. It’s also worth a read.

I blogged in August, at the time of our last Iowa poll, about our methodology there, and we followed the same random digit-dialing procedures this time. Again there’s a lot of winnowing involved in getting down to likely voters: to get 500 likely Democratic caucus-goers we had to interview more than 4,800 adults in Iowa. That’s a lot of calls.

Sampling methodology is a critical point of differentiation among surveys. Another difference is in the number of undecideds – just 3 percent in our survey of Democrats, 4 percent on the Republican side, vs. anywhere from 10 to 16 percent in other recently released Iowa polls.

Undecideds in fact are pretty much of a misnomer, given the construct of the question all these polls ask – in essence, "If the election were today, for whom would you vote?" If the election were today, and you truly were voting (which is what all that “likely voter” winnowing is meant to approximate), then “undecided” would not be an option. Indeed we find that likely voters do have preferences, and with very little encouragement are perfectly willing to share them with us. Thus low undecideds, in our view, represents better pre-election polling.

For sure some people have lightly held preferences, and that’s well worth knowing and analyzing. But those aren’t undecided voters, they’re movable ones. We measure them in this Iowa poll, as elsewhere, by asking people if they might change their minds, and if so, how likely that might be. It’s an important element of the story, underscoring the substantial room to move that still exists in the contest.

The bottom line, as I’ve suggested in a previous blog on election polls, is that we’re certainly not trying today to predict who’ll win the Iowa caucuses in six weeks, nor even to focus singly on the horse race. Our intent is to try to understand the election as it unfolds – in Iowa as elsewhere, the issues and candidate attributes that people care about, the judgments about the candidates they reach, the various groups and competing interests involved. I hope you find it informative.

November 19, 2007 | Permalink | User Comments (31) | TrackBack (0)

Thompson and the NRLC

November 13, 2007 1:39 PM

It’s worth watching whether the National Right-to-Life Committee’s endorsement of Fred Thompson pays off with any votes. Two things we do know are that Rudy Giuliani, naturally, is more vulnerable among anti-abortion Republicans – and that so far they’ve been mightily fractured on an alternative.

Among Republicans who support legal abortion, Giuliani has 39 percent support in our last ABC/Post poll; among anti-abortion Republicans his support declines to 28 percent, vs. John McCain at 18 percent, Thompson at 15 percent, Mike Huckabee at 14 percent and Mitt Romney at 12 percent. If that all were to coalesce, it certainly could add up to hurt Giuliani.

Overall, as noted in our "social issues" analysis last week, 55 percent of Americans support legal abortion – steady the last five years, and almost exactly the 12-year average in ABC/Post polls. That ranges from a high of 78 percent of liberal Democrats to a low of 31 percent of conservative Republicans.

Among all Republicans and Republican-leaning independents, 43 percent say abortion should be legal in all or most cases; 56 percent illegal. The "illegal" number includes 38 percent who say it should be illegal in most cases, and 18 percent who say it should be illegal in all cases.

In a general election match-up, Giuliani leads Clinton by about 2-1 among Americans who say abortion should be illegal in most cases; but, interestingly, he and Clinton roughly split those who say it should be illegal in all cases. Presumably that's because, seeing no meaningful difference between the two on abortion, they're discounting it in their preference.

Equally interestingly, though, is that among these especially strong abortion opponents, Thompson and Clinton (and Romney and Clinton) also are about even. In our poll it’s McCain, not the Right-to-Life Committee’s choice of Fred Thompson, who does best against Clinton among the strongest abortion opponents.

See also Jake Tapper's piece on our politics page.

November 13, 2007 | Permalink | User Comments (2) | TrackBack (0)

Swinging Singles?

November 12, 2007 11:55 AM

It’s the latest group du jour – the population group that’s said to hold the keys to the next election. In past years soccer moms have been paraded by, along with NASCAR dads, security moms and a rumba line of others. At the moment it’s single women, supposedly prowling around for the perfect candidate like a political alter-ego of Carrie Bradshaw on “Sex and the City.”

Peer a bit more closely though, and the view from behind that Cosmopolitan glass might look a little different.

The first thing to know is that “single women” look less like the cast of “Sex” than they look like the supermarket checkout line. All types of unmarried women fall into the category; nearly half of single women are divorced or widowed. Their median age is 48.

Nor are single women a swing group. Within the Democratic Party, Hillary Clinton’s support for the nomination is essentially identical among married and unmarried women alike, 55 and 51 percent, respectively. (Clinton also leads for the nomination among Democratic men in the latest national ABC/Post poll, albeit by less of a margin.)

In a general election contest, single women continue to hold the role of core Democrats. In a match-up against Rudy Giuliani, single women favor Clinton by a 2-1 margin, 65-30 percent. That’s similar to what it was for John Kerry in 2004, Al Gore in 2000 and Bill Clinton in 1996 and 1992. A swing voter group is one whose majority allegiance shifts from party to party in presidential elections. Single women clearly don’t fit the bill.

 

 

A glance at political affiliation underscores that reality. Unmarried women are far more likely than other Americans to identify themselves with the Democratic Party: Two-thirds are Democrats or lean Democratic, dwarfing the party’s advantage in comparable groups. A get-out-the-vote campaign focused on unmarried women necessarily would disproportionately benefit the Democrat in a general election contest. Indeed the activist group “Women’s Voices, Women Vote” identifies unmarried women as “the greatest untapped resources for increasing political support for progressive causes in America.”

                            

While unmarried women disproportionately are Democrats, what “progressive” means, or how it applies, is tougher to tell. As a group they are no likelier than other women (or than unmarried men) to support gay civil unions, legal abortion or a legal-status program for illegal immigrants.

In one difference, unmarried women are the only comparable group in which a majority (53 percent) sees “larger government with more services” as preferable to “smaller government with fewer services.” Income and race may be factors; unmarried women, like unmarried men, include more racial minorities and have lower average household incomes, two groups that also tend to prefer a more active government.

We can, by the way, trim this population down to a closer Carrie Bradshaw lookalike – not single or unmarried women, but “never-married” women. It's a yet more solidly Democratic group – no different in primary preferences, but in a general election match-up Clinton's lead over Giuliani advances to 3-1 among never-married women. On other issues never-marrieds look much the same as all unmarried women, with a few differences related to age, notably on gay civil unions and legal status for illegal aliens.

Ultimately, it’s married women who are the more critical political group. They’re much more likely to be true swing voters. George W. Bush won married women by 55-44 percent in 2004, a key element of his re-election victory. (While they were commonly called “security moms” the presence or absence of kids didn’t actually seem to figure into the equation.) Married women split evenly between Gore and Bush in 2000. Clinton won them in 1996, and ran evenly with them against George H.W. Bush in 1992. This is a group whose preference shifts.

In our latest poll, while Clinton sailed away with unmarried women, married women divided between her and Giuliani by a much closer 50-47 percent. It’s where those married women end up – not their single, Cosmopolitan-sipping counterparts – that’s likelier to play a crucial role in 2008.

November 12, 2007 | Permalink | User Comments (3) | TrackBack (0)

Robertson and Rudy: Does it Matter?

November 07, 2007 11:15 AM

It’s an open question whether Pat Robertson’s endorsement will do more to help Rudy Giuliani’s presidential campaign - or to call into question Robertson’s own credibility among a substantial number of evangelical white Protestants.

Giuliani clearly could use help in this core Republican group. He only runs about evenly with John McCain and Fred Thompson among evangelicals in the latest ABC/Post poll, while holding a 20-point lead among all other Republicans and Republican-leaning independents.

                                                
It may matter: Evangelical white Protestants account for nearly three in 10 leaned Republicans, and their motivation and cohesiveness can make them a dominant force particularly in low-turnout primaries.

Evangelical white Protestants, who predominate in the South, are not a natural affinity group for Giuliani, himself a Catholic from New York. But the real stumbling block is on social policy: Giuliani supports gay civil unions and legal abortion; among Republican evangelicals, 73 percent and 70 percent, respectively, oppose these.

It seems a stretch to suppose that evangelicals would surrender deeply held policy positions on the basis of Robertson’s endorsement, or anyone’s. There does look to be some room for Giuliani: In a poll we did in June, 76 percent of evangelical Republicans said his position on abortion and civil unions made them less likely to support him, but fewer – 47 percent – said there was “no chance” of it. Still, that also means Robertson is backing a candidate whom nearly half of Republican evangelicals have flatly ruled out. (On another issue, last winter 44 percent of Republican evangelicals said they'd be less likely to support a candidate who's been divorced twice, as has Giuliani.)

Perhaps a better question is how much personal endorsements matter at all. Sam Brownback (whose own support for the nomination peaked at 2 percent before he bailed out of the race) lined up behind McCain today. More will follow, all duly hailed by the honored candidate and reported as news. Some endorsements indeed may matter – if, as with some unions, they offer at least the prospect (if not always the reality) of disciplined organizations at the foot-solider level.

As for others, well, we just might keep in mind what Al Gore did for Howard Dean.

See the rest of our coverage of today's endorsements on the Political Radar, on Jake Tapper's blog, and in this piece on Brownback.

November 7, 2007 | Permalink | User Comments (5) | TrackBack (0)