Two things to discuss in this diary that have come up recently:
- SUSA came out with another poll which showed something I most assuredly did not expect, notably one of the three races not acting like the other two. What's most surprising is which one it was: The Lt. Governor's race.
- I feel compelled to talk about the effects of race in the Virginia campaigns this cycle, particularly alongside the comments made about anti-Obama feelings. Race seems to be playing an issue in Virginia politics, even if everyone on the ballot in many areas is white.
First, SUSA
First things first, SUSA's latest poll is a bit of an eyebrow-raiser. Not because of what it shows in any one of the three races, but because of what it shows in the Lt. Governor's race vis-a-vis the other two. Notably, in the last umpteen polls (including the immediately previous one which I covered in my last diary), the difference between the three statewide races has been mostly negligible...one would be +6, the others would be within 2-3 points of this, mostly affected by varying numbers of undecideds and minor quirks in crossovers. Some weren't just sampling noise, they were effectively rounding errors. In the last poll, for example, SUSA showed the following:
Governor: McDonnell 55, Deeds 41 (R+14)
Lt. Governor: Bolling 54, Wagner 41 (R+13)
Attorney General: Cuccinelli 53, Shannon 42 (R+11)
The latest poll was a little different, however:
Governor: McDonnell 54, Deeds 43 (R+11)
Lt. Governor: Bolling 57, Wagner 40 (R+17)
Attorney General: Cuccinelli 43, Shannon 43 (R+10)
What is key here is not that Bolling gained three points. That's noise. It's that he gained while McDonnell fell, and that Wagner fell while Deeds rose. This hasn't been happening thus far in the campaign.
My best guess is that the shift among women to Deeds that SUSA made a big deal about (three guesses as to the most plausible culprit for that, and the first two don't count) simply didn't translate in the Bolling race like it did elsewhere. Bolling built up a lot more ground among men than he lost among women, while the opposite was true in the other races and increases among men just didn't offset losses among women.
The "why" of this is the more puzzling question: Why did Bolling fare much better than the others, particularly given how low-profile the race has been and how closely he's tied himself to McDonnell? My best guess is that Bolling's cash edge over Wagner, roughly by 2:1 (and possibly a bit more than that, depending on how September went for the two fundraising-wise), combined with a lack of missteps by Bolling to let him weather things a bit better. Other than that, though, I'm really at a loss on this one, as Bolling hasn't really done anything to merit such a swing versus McDonnell...and Wagner hasn't done anything that I've heard to merit such a drop versus Deeds.
Of course, it's also possible that this is your classic rogue poll, and it just included a lot of people who really don't like Wagner for some odd reason. Time will tell.
Race in this Race
One thing that is always worth thinking about in Virginia politics how much of a case of MPD the state has at times: NOVA may as well be Maryland, New Jersey, or Connecticut. Southwest Virginia is remarkably similar to parts of Kentucky and West Virginia. Most of the rest of the state acts a lot like most of the South.
A lot of this played out in the 2008 presidential primaries: Clinton blew the doors off of Obama in Southwest Virginia, but didn't even break 40% in all but two counties east of I-95. The reason for this is two-fold:
- Clinton could count on the white vote in many areas in the south due to racial issues that have kept cropping up at an irritating rate over the last year and change.
- Obama could count on the black vote in the primary coming to him by between 10:1 and 20:1. The key is that he was also able to tap many of the same groups that backed Jerry Brown in '92 on the left wing of the party to augment them (if you look at primary results in New England, Hillary won the Tsongas states and Obama won the Brown states). The ultimate key seems to have been Obama's ability to tack on a large part of the black vote and a few more liberals to Brown's coalition in '92.
All of this is highly relevant in Virginia's race. In Southwest Virginia, where there are neither many blacks nor many liberals, Obama got squashed. Defining the region momentarily as the 276 Area Code, John Kerry broke 40% in about 8 counties and won two counties. Obama managed 40% this in only four, and won zero.* In one county, Tazwell County, McCain added roughly 8 points onto Bush's showing in 2004 (going from 57% to 65%) while he was under-performing Bush by 7% statewide.
In the most recent PPP polling, Deeds is getting a whopping 25% in this region to McDonnell's 70%. That's not a loss, that's a massacre, and this is in an area that Deeds got about 42-43% in last time around (I can't put together an exact number right now, but he won several counties while breaking 45% in most of the others). Clearly there's something in the water out there, as even accounting for a weaker overall performance Deeds is off about 8-10% there.
My best guess is that the racial factors that led to Obama getting trashed in Appalachia are at play again in this race. In 2005, Kaine cracked 40% of the white vote in most polls, and was winning it in the last SUSA poll before the election 48-47. By contrast, the three most recent polls for which I have crosstabs (SUSA, SUSA, and PPP) all show Deeds stuck in the mid-30s while McDonnell floats along over 50% (a margin of 15-20 points for McDonnell).
I'm willing to posit that this problem probably doesn't carry into NOVA, but in a lot of the rural areas of Virginia. My basis is that when PPP did their poll on Obama's birth records back in August, they gave us a present in the form of their crosstabs. In the 540, 434, and 276 area codes (primarily made up of the Shenandoah Valley, far SW Virginia, and "Old Southside" respectively), "No" got 32%, 26%, and 29%...the three highest results statewide. 804 (Central VA, incl. Richmond but also a lot of rural counties stretching east to the Chesapeake Bay) also yielded a decent number, while 757 (Hampton Roads) and 703 (NOVA) yielded a far lower share (though 757 also generated a decent block of undecideds).
This problem is mostly rural/small town and heavily Republican...but it isn't exclusive to Republicans, and has probably messed with party ID a bit in these areas, too. Basically, though, issues tied to Obama among white voters do seem to be trickling down onto Democrats in general and Deeds in particular. This is probably very closely connected to Deeds' unwillingness to tie himself to Obama (the debate question on this leaps to mind).
The problem is, almost assuredly, not just race itself...but race is playing a factor somewhere in the mix. In light of the issues that Obama had in Appalachia, I think that he lost a good chunk of the vote in areas like SW Virginia over race...but much like Lyndon Johnson's decision to drop the white vote in the South in the '60s, I suspect that this trade will help the Democrats in the long term nationally. Unfortunately, absent the thumping black turnout Obama got plus the extreme frustration/anger felt at the GOP in 2008, this trade does present problems for Democrats in Virginia this time around.
Outside of NOVA and most of Hampton Roads, race is still an issue in politics. Obama's nomination seems to have broken many voters in these areas of any Democratic tendencies, and if this is the case then it is a shame that so many people would ditch their party over the race of one candidate...particularly when many of them voted for Doug Wilder in 1989.
However, based on evidence from the presidential primary last year, the general election last year, and the 2005 and 2004 elections I am left with race as the most obvious culprit, and thus my conclusion is that regardless of the candidate, there are some people who are going to hold Obama's race against the Democratic Party as a whole.
*Note that this excludes cities, which aren't part of counties due to Virginia's odd habit of making cities mostly independent of the surrounding county. On most election maps, cities are virtually invisible, making a quick count very hard.