Usual disclaimer: I support Chris Van Hollen. But I submit that a poll is either good or it’s not. You decide.
A new poll for the Washington Post, conducted in association with the University of Maryland, shows Donna Edwards with a 44-40 lead over Chris Van Hollen. There are no geographic cross tabs, but there’s an interesting chart showing racial and gender splits.
In the body of the article, there’s some numbers on the racial divide even among women. White women support Van Hollen by a margin of 23 points, while African-American women back Edwards by 51 points.
There’s some data here, but it’s quite frankly almost useless, focusing on questions of likeability and “who do you think cares most about you?” stuff. Even the racial and gender splits highlighted in the Post article are nowhere to be found in the data. It’s possible that the Post is holding back some of the juicy stuff, as there are questions listed as “held for release.” We’ll have to wait and see what else they release.
On the ever-present question of the racial composition of the actual voter pool, the poll has a very heavy tilt in favor of Edwards: the black percentage of the predicted electorate is at the highest end of all the polls so far (45%), matching the Goucher Poll number from two months ago. Y’all know what I thought of that poll.
Perpetual reminder – with Barack Obama on the ballot in 2008, African-American voters constituted 37% of the electorate. I don’t know what the white percentage is for the current Post poll, but we know there are some Hispanic votes as well around the state. Even if “other” is only 6-8% here, that means white voters will constitute less than half of all votes statewide (47-49%). That. Will. Not. Happen. This is a problem for this poll, as it has been for virtually all of them.
We’re close enough to the election now to make some predictions about what’s going to matter in this race. In no particular order:
1. Baltimore. Edwards is going to win the City, while the suburban counties are going for Van Hollen. Can either candidate cut into the other’s numbers here?
2. Turnout. Answering a phone call and saying you’re sure you’ll be voting is not the same as actually voting. Who does a better job of turning out their base areas and constituencies will have a huge effect. Van Hollen has a better turnout operation than does Edwards, and by virtue of his endorsements should have the benefit of the networks of state and local officials around the state. Will it have an impact?
3. TV. Van Hollen has a lot of money in the bank. Edwards doesn’t. Even with Emily’s List on her side, outside groups pay a heavy premium for air time as compared to candidates themselves. A dollar of IE spending is worth less than a dollar of candidate spending.
4. Tone. So far, the conversation has been pretty polite, although in recent face to face events between the candidates, the rancor is rising. Will it stay where it is or will it escalate? Will one or the other candidate go negative in advertising, particularly TV?
5. SouthEastWest Maryland. If this race stays as close as it might, it could well be decided by Southern Maryland (Charles, St. Mary’s and Calvert), Western Maryland (Frederick, Washington, Allegany and Garrett), or the nine counties of the Eastern Shore. Overall, these areas should break heavily for Van Hollen, but that will only matter if the race is close, because there are vastly more votes in the Baltimore region and the DC suburbs.
Summary: most of the intangibles, I submit, work in Van Hollen’s favor. But the race has to be close enough for them to matter.
And hey, media guys – hire an exit polling firm and get us some Election Day data, wouldjaplease? Thanks.