| All three Democrats -- Creigh Deeds, Jody Wagner and Steve Shannon -- trail their Republican counterparts by roughly 6%, according to a PPP poll out yesterday. McDonnell leads Deeds 49% to 43%; Bill Bolling leads Jody Wagner 46% to 40%; and Ken Cuccinelli leads Steve Shannon 45% to 38%. The margin or error is 4.0%.
Looking at the numbers in detail, I am struck by four things that give me hope, and two facts that give me pause.
1. When the margins numbers are uniformly 6% to 7%, that tells us that the respondents are going on name recognition and Party ID only. The fact that Cuccinelli is getting 45% of the vote tells me that 45% of the respondents are pretty favorably inclined to push the Republican button; it does NOT tell me that Cuccinelli has 45% of the people who know him and like him. Not in July.
2. The breakdown on the poll includes 34% Democrats, 33% Republicans and 33% other (presumably independent). Last year, the polls assumed that Dems would outnumber Republicans by about 38% to 30% (the final Survey USA poll was typical). The actual exit poll data shows that Democrats outnumbered Republicans among voters by 39% to 33%. Last year's Presidential poll also assumed that the liberal-moderate-conservative split would be 16%-46%-31%, whereas this PPP poll assumes an electorate of 19%-42%-38%. This poll sample is also much older than the expected electorate last fall -- this poll assumes 8% of the voters will be under 25, as opposed to 25% in last fall's polling. Of the electorate that came to the polls last November, 12% were under 25. The PPP Poll sample was 22% over 65; the actual electorate last fall was 11% over 65. See here for last November's exit poll data. In other words, this poll sample is more Republican, older and more conservative than the poll samples last year, and more so than the electorate that actually come to vote last fall. If Virginia Dems can duplicate the 2008 electorate, they will sweep the races. Of course, that's a big "if", but it gives Dems a target to shoot for.
3. The PPP poll shows the Republicans consistently getting 15% to 16% of the African-American vote. While we should be cautious about assuming anything, it seems most likely that they will return to their usual pattern of giving Democrats about 90% of their votes.
4. In 2005, Tim Kaine won the 703 area code (Northern Virginia) by roughly 61% to 39%. Creigh won the same area in 2005 by 58% to 42%. The PPP poll shows the 703 area code going for Creigh Deeds by 53% to 38% -- roughly his 2005 margin. There is room for improvement there, though it will be tough.
Those are the optimistic thoughts, after analyzing the poll.
I have two pessimistic thoughts:
a. We know we have to get some federal voters to the polls this time. That is always hard, but it will be particularly hard this year. The state electorate tends to be older and more conservative, but if we can bring out some of the Obama voters, we'll be OK.
b. The fact that 45% of the electorate starts with at least a strong inkling that they want to vote for the Republicans. This is a more important figure than the responses given when poll respondents are asked whether they belonged to a particular party.
Conclusion -- the poll sample is flawed, the poll shows that we are within striking distance in all three races, and the flaws, if rectified in the final analysis, will work in our favor. |
