Quinnipiac released a poll that shows George Allen leading Tim Kaine 44% to 42% in Virginia’s US Senate race. Other candidates on either side were not polled, but there was an option for the generic “someone else” that totaled 2%, combining alternate candidates on both sides.
There are some other interesting bits in this poll, but first, a few words on these polls in general.
This Quinnipiac was weighted. As I discussed last week, weighting a poll is a statistical adjustment that into account the makeup of the polling area, in this case the state of Virginia, versus the makeup of the polling sample. In other words, if you call a greater percentage of Republicans than actually exists in the state, then to get an accurate picture, you need to adjust the poll to account for that.
In last week’s poll, Public Policy Polling, a Democratic polling outfit sampled more Democrats than Republicans by a good bit. More than the current ratio of Republicans and Democrats in Virginia. My problem with the PPP poll is that they released the raw, unweighted results that showed Tim Kaine with a 5 point lead. So, I weighted the poll, like an honest pollster would do and the result was a 1 point Allen lead.
The Quinnipiac poll properly adjusted the results in their release and it shows, basically, the same thing as the PPP poll when properly weighted. The race is pretty much even, which was the same conclusion I reached in the weighted PPP poll.
One clue that the PPP poll was suspicious was their inclusion of a statement by a “respected” statistical wonk named Nate Silver. If you don’t know the name, then you obviously don’t read far left media outlets like DailyKOS or the New York Times. But the inclusion of such a statement tells me that they know their poll is blatantly partisan. Now I don’t make a claim that the poll itself was flawed. It’s all in the presentation. I believe that PPP was trying to show a post AP debate bounce for Kaine, and it simply is not true. The debate seems to have had no effect at all for Allen or Kaine.
And one trend I have noticed about some of the partisan polls. They tend to publish partisan conclusions are an attempt to sway opinion rather than simply measure public opinion. Most people are aware of the trick questions and how some people use “push polling” to influence the answers. Those tricks are usually easy to spot, so most “respectable” pollsters don’t use them. But the devil is in the interpretation.
And we all know that the saying goes “polls tighten as election day approaches”. And this is absolutely true. Most pollsters want you to think that this is a result of people finally making up their minds. And that is part of it. But mostly, the pollsters understand that when the election is over, they will be judged on the accuracy of their last poll prior to the vote. The closer they are to the actual outcome, the more “accurate” they will be perceived. So the last few days, they all take great pains to carefully weight the results and you will also see that the results usually move from “All adults over 18 – registered or not” to “Registered Voters” and finally to “Likely Voters”. The last category, Likely Voters, is by far the most accurate. But it takes a lot more calls (and money) to weed out the less than likely voters, so they all go for the low hanging fruit until the last few weeks of the campaign.
But there is always some really good information hidden – and ignored – in the answers. And this Quinnipiac poll has a few.
- Newt Gingrich is favored over Romney by Republicans, but loses to Obama.
- Romney leads Obama in a head to head by 2 points.
- Republicans are more enthusiastic about this election than 2008 than Democrats (42% – 29%)
Allen – Kaine Senate Election
- Kaine leads Allen in support from people making less than $30K (47% – 38%) which is closer than most would expect.
- Allen leads Kaine in every other income category ($30K – $50K, $50K – $100K and $100K +) which is significant.
- Allen’s support has remained steady around 44% to 45% for the past 90 days while Kaine has dropped to 42% indicating a slight downward trend for the Democrat.
- The favorable rating of both candidates dropped 4-5 points since October and are about even.
- The unfavorable rating of both candidates improved a couple of points and are still about even.
- The number of people who don’t know enough about either candidate rose several points since October and are about even.
Overall, things have remained pretty even between George Allen and Tim Kaine. There is probably a bit more cheer in the poll for the Allen Camp than the Kaine camp in this news.
What the polls say is that Allen had broad support across all income levels, which is a positive thing. And while most Virginians remember Tim Kaine as the most recent Governor (prior to Bob McDonnell, the current Governor) and his highly visible job as Democratic National Committee Chairman and main Obama cheerleader, George Allen has been out of politics for almost 6 years. To be even with Kaine, who has been on the news and airwaves constantly for those same 6 years, has to give a great deal of encouragement to Allen. Based on exposure the last 6 years alone, Tim Kaine should have an advantage. But that gap has closed.
This poll is also the first taken since the AP debate, and while the movement was a slight shift – within the margin of error – towards Allen, it is accurate to say that debate was more helpful to Allen then to Kaine.
The fact that the number of people who don’t feel they know enough about either candidate has gone up is also important to note. It tells us that people are still forming an opinion and that there is much work to be done by both campaigns to get their message out there.
Tim Kaine has a difficult task ahead trying to convince people that his support of unpopular programs like ObamaCare was the right move, and George Allen needs to establish communications with younger voters (under 35) where Kaine has an advantage. Allen should be pleased that those 35+ give him more favorable ratings than Kaine because those of that age remember Allen as both Governor and Senator and view him more positive than they do Kaine.
While this poll is statistically even, Allen looks to be trending positive and Kaine is steady to losing ground.