In the battle of the spouses the early edge is Michelle Obama's, in favorable views and intensity of sentiment alike. But there are sharp differences among groups, and plenty of room to move for the less well-known Cindy McCain.
Forty-eight percent of Americans in a new ABC News/Washington Post poll see Obama favorably, vs. 39 percent for McCain, a 9-point Obama advantage. Slightly more, though, also view Obama unfavorably – 29 percent vs. McCain's 25 percent.
Substantially more, 36 percent, haven't yet formed an opinion of McCain, vs. 23 percent in Obama's case. For both, those are sizable numbers who've yet to make a judgment.
The popularity of presidential candidates' spouses does not drive vote preferences. But in contests where every advantage can count, spouses do play a very public role. Cindy McCain is highlighting her support for children's charities with a visit to Vietnam this week, while Michelle Obama hosts the ABC program "The View" on Wednesday.
DIFFS – There are big differences among groups in views of the two women, mainly driven by political partisanship. Obama's favorable score is 14 points higher among women than McCain's, 54 percent vs. 40 percent; as in many of Obama's other best groups, the chief reason is simply because women are more apt to be Democrats.
An even more striking gap may cut to Obama's independent persona; among the two in 10 Americans who call themselves feminists (men and women alike), 60 percent view her favorably. That drops to 45 percent among non-feminists – who are twice as apt as feminists to see her unfavorably.
Obama's ratings peak at 84 percent favorable among African-Americans, 66 percent among liberals and Democrats alike and 61 percent among young adults, age 18-29. Not surprisingly, those are among her husband's core groups; indeed it's his support that seems largely to drive views of his wife. Among people who prefer Barack Obama for president vs. John McCain, 73 percent like Obama's wife, too.
McCain's support, naturally, inclines the other way. Though the difference is less striking, she's better rated by non-feminists (41 percent favorable) than by feminists (33 percent). She does best with Republicans (62 percent favorable) and with her husband's supporters (56 percent of whom like her, too). But her favorability rating among conservatives (46 percent) is a full 20 points below Obama's among liberals.
Obama also inspires stronger opinions: The number of Americans who have a "strongly" favorable opinion of her is double the number who say so about McCain, 21 percent vs. 10 percent. On the negative side, 18 percent see Obama strongly unfavorably, as do 12 percent for McCain.
In only a few groups does either of these women's unfavorable ratings outscore their favorables. More African-Americans, under 30s and Democrats see McCain unfavorably than favorably (it's close among liberals); and more Republicans and conservatives see Obama unfavorably than favorably (with seniors about evenly split).
Finally, in a hangover from the hard-fought primary campaign, Obama's favorable rating is a good deal higher from her husband's primary supporters – 74 percent – than from Hillary Clinton's, 51 percent.
VOTE – Again, views of the candidates' spouses do not significantly influence vote choices. In a regression analysis – a way of sorting out the independent influence of various factors – support for Barack Obama is predicted by partisanship, preference for him over John McCain on personal attributes, preference for Barack Obama to handle top issues and favorable views of him – but not by favorable views of his wife. The same holds true for Cindy McCain in predicting support for her husband's candidacy.
A result from an ABC/Post poll in December pointed the same way: Sixteen percent said that in deciding their vote they give a "great amount" of weight to the candidates' spouses; by contrast 65 percent put a great amount of weight on the issues, 58 percent on the candidates' personal qualities and 54 percent on their professional ability.
METHODOLOGY - This ABC News/Washington Post poll was conducted by telephone June 12-15, 2008, among a random national sample of 1,125 adults, including an oversample of African Americans (weighted to their correct share of the national population), for a total of 201 black respondents. The results from the full survey have a 3-point error margin. Sampling, data collection and tabulation by TNS of Horsham, PA.