Pollster.com

Articles and Analysis

 

Odds and ends

Topics: 2008 , Barack Obama , Exit Polls , Hillary Clinton , John McCain , National Journal , NY Times , Pew Research Center , Quinnipiac

Here are three interesting poll-related reads I encountered over the weekend, all dealing with the Obama-Clinton race: .

First, my National Journal colleague Ron Brownstein looks closely at what recent surveys have to say about the potential general election coalitions for Barack Obama and Hillary Clinton, and concludes that while "Obama offers greater potential rewards as a nominee," he "also presents greater risks:"

In a recent Pew Research Center survey, for instance, Obama carried independents against McCain by 6 percentage points, while McCain carried them against Clinton by the same amount; the difference mostly reflected Obama's stronger showing among independents earning at least $50,000 annually. Other surveys, such as a Quinnipiac University poll in the key battleground of Pennsylvania, have found that Obama also swipes more Republicans from McCain than Clinton does.

This all tracks Obama strengths familiar from the primaries. But primary-season trends more troubling for Obama are also persisting. In the national Pew survey, and in Quinnipiac polls of Ohio and Pennsylvania, Obama lost more Democrats to McCain than Clinton did. In the Pew survey, Obama struggled particularly among the same blue-collar white Democrats resisting him in the primaries: Fully 30 percent of white Democrats earning less than $30,000 a year preferred McCain over Obama. Clinton would lose only half as many of them to McCain, the polls indicate. In the Quinnipiac surveys, Clinton likewise outpolled Obama against McCain among white women without college degrees, a key general election swing group that has overwhelmingly preferred her in the primaries.

Second, Matt Bai devotes his New York Times Magazine column to an argument that has raged in the blogosphere for many weeks: "Obama wins in major urban areas but can’t seem to win in urbanized states, while Clinton wins in rural communities but consistently loses in rural states." To illustrate the point, the Times created this intriguing chart showing exit poll results in six states tabulated across urban, suburban and rural areas:

03-17 NYTimes graphic

He reviews several relevant theories for the pattern, but seems most taken by this one:

What this suggests, perhaps, is that living in close proximity to other races — sharing industries and schools and sports arenas — actually makes Americans less sanguine about racial harmony rather than more so. The growing counties an hour’s drive from Cleveland and St. Louis are filled with white voters whose parents fled the industrial cities of their youth before a wave of African-Americans and for whom social friction and economic competition, especially in an age of declining opportunity, are as much a part of daily life as traffic and mortgage payments. As Erica Goode wrote in these pages last year, Robert Putnam and other sociologists have, in fact, found that people living in more diverse areas evince less trust for others — no matter what their race. Maybe it shouldn’t surprise us that while white Democrats in rural states are apparently willing to accept the notion of a racially transcendent candidate, those living in the shadow of postindustrial atrophy seem to have a harder time detaching from enduring stereotypes, and they may be less optimistic that the country as a whole would actually elect a black candidate.

Third, on the Monkey Cage, GWU political scientist Lee Sigelman blogs a compilation the various "opponent dissatisfaction" scores from the exit polls since South Carolina. In other words, what percentage of Clinton voters say they would be dissatisfied if Obama is the nominee, and vice versa.

03-17 sigelman dissatisfaction.png

Sigelman notes several patterns in these numbers, but this one addresses the question I have heard most often:

Since Super Tuesday (February 5), Obama’s supporters have expressed greater dissatisfaction about a potential Clinton nomination than they were doing before then — again, presumably as a result of Clinton’s attacks on their favored candidate. For Clinton’s supporters, there has been little movement over time in their dissatisfaction with Obama’s possible nomination.

 

Comments
Andrew S. in California:

Its interesting to see the Opponent Dissatisfaction Index. I can see both candidates becoming more polarized as the race goes over time.

____________________

Ciccina:

I read Matt Bai's column and it left me cold. The rationale excerpted here seems particularly off base.

I don't know why he would make the leap from:

"Obama wins in major urban areas but can’t seem to win in urbanized states, while Clinton wins in rural communities but consistently loses in rural states..."

to:

"Maybe it shouldn’t surprise us that while white Democrats in rural states are apparently willing to accept the notion of a racially transcendent candidate, those living in the shadow of postindustrial atrophy seem to have a harder time detaching from enduring stereotypes and they may be less optimistic that the country as a whole would actually elect a black candidate."

Ever since the New Hampshire results came in, there's been a subtle and not-so-subtle meme floating around that white voters who don't support Obama are doing so because they are racially intolerant. But where's the proof?

Bai's argument only makes sense if one presumes that Obama is such a manifestly superior candidate, the only reason someone could oppose him would be racism. Which is clearly not the opinion of a great many people.

Bai doesn't consider a reciprocal argument: that voters in traditionalist red states are not comfortable voting for a woman to be our national pater familius and don't believe a woman can win, while blue state working class voters, who are more likely to be part of dual-income or single-income female-headed households, have transcended gender bias. Some of the red states - some of the blue ones that Obama won, too - have never elected a woman to Governor and/or Senate, despite the fact that 51-52% of that state's population is female.

For another, Bai doesn't give any evidence, even anecdotal, that racial intolerance is a more salient issue for white working class voters in states like Ohio than the economy and other bread and butter issues. Hillary is well known as an intelligent, hard worker who has detailed plans for the future. By contrast, Barack is the very image of the snooty, elite, esoteric pol. There are probably a dozen reasons more likely than race to be a decisive factor for working class voters.

What Bai posited was so insulting to so many that I have to wonder if he really thought it through (he seems way too intelligent to make this kind of generalization, not without some kind of evidence).

____________________

I was most taken by the Obama carrying greater risks and greater potential rewards. What disturbed me, however, was the lack of age data in the crosstabs (please tell me if I missed it). For example, I would think that non-college women is greatly confounded by being Baby-Boomers or older for that sample.

My thesis is that generational identity is the most important factor in this election, but I have spotty data at best to back it up.

____________________

Joseph E:

To reply to Ciccina Re:

"Bai doesn't consider a reciprocal argument: that voters in traditionalist red states are not comfortable voting for a woman to be our national pater familius and don't believe a woman can win, while blue state working class voters, who are more likely to be part of dual-income or single-income female-headed households, have transcended gender bias."

I don't see how the data could fit your counterargument. Bai was talking in particular about cities versus rural areas (suburbs seeming to vote somewhere in between). In the 6 states listed, Obama won urban areas (cities >50,000) in all but California (where the Latino vote likely won the cities for Clinton), even in Tennessee and Ohio. Clinton did not always win the rural areas, but she did carry them in Missouri despite narrowly losing the state, and generally won big in the rural areas if she won at all.

Now, if sexism were the reason for Obama's wins, then why would he do better in the big cities than in the country? As a country boy from a very small Northwestern town who now lives in a Southern California urban area, I can confirm that the country was much less progressive on gender than the city, in general. From my visits to Houston and rural Texas, Miami and rural Florida, and so on, I think this holds true in the south.

So why would rural voters in Ohio pick Clinton, while the cities chose Obama? Again, why do Missouri suburbs vote Obama, while the rural counties of the state went overwhelmingly for Clinton? Looking at the pattern, it seems like racism may win out over sexism, at least among white rural voters. I have trouble imagining that the cities of Ohio are more misogynistic than the small towns.

I'm sure that sexism is hurting Clinton among some male (and female) voters (though most of those, fortunately, are Republicans, and aren't voting right now anyway... if we ignore the Limbaugh cross-over voters), but her gender also seems to have helped her among women in many states. My wife was on the fence, and in the end went with Hillary for the chance of electing a woman as president.

But similarly, there seems to be a subtle racism against Obama. Ignoring the clear racially divided votes in Mississippi and other Southern states, there is a still a trend against Obama in many white populations, confirmed by multi-variant analysis:
1) Areas with Southern Baptists (in other words, white protestant Southerners)
2) Lower educational attainment
3) Lower income
4) Greater age
5) Rural areas
Don't those sound like the folks who might be wary of Obama due to his name and the color of his skin?

I'm not saying they are outright bigots (these are Democrats we are talking about), except for some of the southerners. But the influence is subtle. I still feel my heart beat faster when I see a group of young black men walking on the street in my neighborhood, a sensation I would never feel about a similar group of vietnamese men (the other big group in this neighborhood), or even a bunch of latinos.

The kicker: the percentage of black population in a state is also an independent variable. Yup, Obama's support goes down (among non-black voters) as the percentage of blacks goes up. This correlates with a roughly northwest to southeast geographical trend. Iowa (2.5% and a caucus, but with high turn-out) Washington (4% black) and Wisconsin (6% black) for Obama vs. Tennessee (17%), Ohio (12%), New Jersey (%15) for Clinton. It seems that living near a black inner city neighborhood or in a former slave state is correlated with fewer whites voting for "the black guy." Out in the Northwest, the Mountain states and the Prairies, race isn't such an issue (in the Democratic party, at least), or perhaps Obama's background is a positive in a state like Washington or Wisconsin.

____________________

Joseph E:

California. Why were we one of the only states where the big cities voted for Clinton? The NYT chart is a surprise; Rural, Suburban and Urban areas voted about the same in California, it seems. Why should this be so?

I think it is fairly east to see why city and country would be united for Obama in Georgia, Wisconsin, and other states in the Northwest or Southeast quadrants of the nation. The old Confederacy has shown a strong racial divide, but the high percentage of black Democratic voters has managed to tip the election to Obama in the rural areas (balancing out the rural white vote). In Wisconsin, and presumably in other states from the Prairies to the Mountains, the cities are not so very different than largely rural state as a whole.

Not so in California. Having lived in a very small town, a very big city, a coastal suburb and a college town, I can say that California is incredibly diverse. The rural folks back north were largely Republican, but with a libertarian streak, and would be happy to send my current latino neighbors "back to Mexico" (no matter their country of origin ;-) ). The Bay Area prides itself on embarrassing the national party every year with displays of craziness; unilaterally performing gay marriages, denouncing Israel, shutting down traffic with bike demonstrations. Southern California still looks a lot more like the "land of Regan", with military families and Suburban sensibilities in all but the most urban areas. And you can't forget that huge latino vote down here.

So what happened? I think the unity between rural and urban areas is just luck. San Francisco vs San Diego; rural mountain towns vs Hispanic farm communities; latte-sipping professional suburbs vs military/industrial suburbs. In the end, Southern California voted more like Arizona and Texas, with the Latino vote making the difference, both in the rural and urban areas.

____________________

Ciccina:

Joseph E, you may be the lucky recipient of my longest post to date ;-)

I agree with much of what you are saying. I don't think the particular sexism argument I put forward is a good one; I only meant that if you are predisposed to see a particular variable as important (racism, or sexism, or perhaps age) you could draw out a narrative that would, on the surface, confirm that presumption without too much difficulty. My point (which I conveyed really poorly) is that focusing on race as a determining factor is interesting, but it ultimately fails to convince because it tries to weave together too many disparate elements, excludes relevant variables and relies on some big assumptions.

I would respond to Bai's original point -- "Obama wins in major urban areas but can't seem to win in urbanized states, while Clinton wins in rural communities but consistently loses in rural states." -- thus: Obama has a massive advantage over Clinton among African American voters. He wins urban areas that have a large Af Am population. The Northern / blue states have such areas, but in addition they have a large population of white and poss. Latino Dem and swing voters; so the Af Am share of the Dem primary vote is less and he doesn't get enough votes overall to put him over the top. In Southern red states, the urban Af Am population votes Dem, but in large part the white vote is Republican / conservative. Since they (mostly) don't participate in the Dem primary, Af Americans make up a proportionally larger share of the primary vote than in Northern states, which translates into an advantage over Hillary.

I wonder if the numbers bear this out.

You write: "there seems to be a subtle racism against Obama. Ignoring the clear racially divided votes in Mississippi and other Southern states, there is a still a trend against Obama in many white populations, confirmed by multi-variant analysis:

1) Areas with Southern Baptists (in other words, white protestant Southerners)
2) Lower educational attainment
3) Lower income
4) Greater age
5) Rural areas

Don't those sound like the folks who might be wary of Obama due to his name and the color of his skin? "

My response is that there are a number of reasons why white voters who fit into these categories - or at least the white female portion of these categories - would be voting FOR Hillary based on issues and character, rather than voting AGAINST Obama because of race. The following are quotes from an excellent 2006 essay by Anna Greenberg:

Re: Greater age: "While unmarried women are close to becoming a base Democratic group, older women remain a swing group. Like unmarried women, older women represent a sizeable chunk of voters - 27 percent - and will be even larger as the Baby Boomers age. They are "crosspressured" by significant economic concerns surrounding health care and retirement security that conflict with socially conservative views, particularly if they are white and blue collar.

In 2004, Kerry won older women by 3 points, 51 to 48 percent, but lost older white women by 5 points - a 4-point drop in support for the Democratic candidate since 2000. In fact, Bush posted the best performance by a Republican among older women since 1988."

[snip]

"There is an important lesson to learn from 2004: when we fail to address their issues, we lose older women voters to cultural conservatism. Even in a campaign whose broad themes are about security, it is possible to target older women voters with messages that address their economic concerns, as Kerry did with his early battleground-state advertising. Unlike younger, socially conservative women, older women want to vote their economic security."

To voters such as these, the choice is between McCain (social conservative) or Clinton (economy, health care, family issues). Obama is almost irrelevant - per the Pew research, he does poorly on these issues and is widely seen as inexperienced. On top of that, figure in character: Hillary rejected divorce. Obama, on the other hand, admits to having used drugs - which totally undermines a parent's message to her own kids that using drugs is a sure path to ruination.

Re: Lower education, lower income voters, Southern Baptist and rural voters:

"White, socially conservative women. Democrats lose white blue-collar women and white married women by large margins; in fact, these women could almost be considered Republican base voters. Yet, there are important openings with these conservative voters. They have concerns about their families that can be addressed by progressives, such as the prevalence of violence and sexual content on television, video games, and the Internet. They are pragmatic and want to protect their children, by making sure they have access to comprehensive sex education in school, and their parents, through stem cell research into chronic medical conditions. Progressives should be able stop the hemorrhage among these women by reframing what it means to care about children and families."

This is another area where its Hillary vs. McCain, with Obama kind of irrelevent. Not only is Hillary seen as more responsive to bread and butter family issues (economy, health care), she's also done things like call video game makers to task for the lurid content of their product, she held her marriage together, and so on. The other choice is McCain, with his social conservative profile. Obama preaches "change" - the very thing high income Dems perceive as "progress," and social conservatives perceive as "risk."

http://www.greenbergresearch.com/index.php?ID=1663

I know the story thus far is that Obama does better among Republicans and independents, but I wonder how accurate this is. If Hillary is (a) pulling out larger-than-before numbers of certain Dem white female subsets, and (b) pulling female swing / undeclared voters into the Democratic column, she would not only be adding white voters to the overall pool (offsetting Af Am impact) but adding voters who would be coming out for reasons that have nothing to do with race.

I don't know if that actually made sense - I'd love to hear your opinion. But at a minimum I think one needs to dive further into the gender, age, economic status, issue preferences and partisan profile of the white vote before any single factor, such as racial tolerance, can be isolated.

------------------

I certainly agree with your assessment of California, having worked on three different campaigns there located in San Francisco, Salinas and North Hollywood/Burbank. Like three different planets. ;0

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR