PPP Virginia: Obama Still Ahead, 50-47

[J. J.]

If it more than +1, he is.

[J. J.]

If it more than +1, he is.

By "he," you mean Obama, and by "what," you mean still ahead?  Why +1?  (Unless you mean up in the poll...where actually he needs to just be up one respondent, even if that rounds to a tie.)

PPP has a slight D bias.  +2 or better, Obama is still holding VA.  +1, and it is a tossup.

[J. J.]

You must be logged in to read this quote.

+3 right now? I'll take it. Better than I expected actually.

Agreed.  They said it has been the closest this year.  I do want to see the internals.

[J. J.]

If the right-leaning snap polls by Gravis, Rasmussen, and We Ask America soon after the debate were valid, then  the PPP poll in Virginia (PPP is as good as any except perhaps Selzer) suggests that the Romney campaign is in deep trouble.

The bump that President Obama got for offing Osama bin Laden was also short-lived. 

You make an assumption, that the race will be static.  That might not be the case.

[J. J.]

PPP has a slight D bias.

No it doesn't.  Nate Silver demonstrated how it actually has a slight R bias.

Well to be fair - and it's not often I'll leap to J.J.'s defense! - pretty much every pollster had an R bias compared to the final result in 2010. PPP less so - that either means they're more accurate or biased towards Democrats, although personally PPP's one of the few pollsters I pay much attention to.

It is one that I pay attention to, but I was thinking about 2008.  It slightly overestimated Obama, but not by a huge amount.  I think it has something to do with their modeling, but it isn't great and it is well within their MOE.  It is consistent with about a one point lean to the Democrats.

I give Rasmussen a higher bias to the GOP, but again, not by a lot.

Frankly, those are the two best pollsters that publish routinely that are out there.  Mason Dixon also tends to be very good, but they don't publish as often.

My rule of thumb is basically, actual result is the race is actually Rasmussen R -2, PPP D -1.

[J. J.]

Well, it's PPP, so that's still more like Obama+1... Still good news on balance. Virginia is slightly more republican than the nation as a whole, so it might be a sign that Romney's bounce didn't make the race a pure tossup. I guess this would translate into Obama+2 nationwide, which is "dangerous territory" but not quite a dead heat.

I'd say Obama +2.

[J. J.]

Again Party ID is just an attitude, not a demographic.

Speaking of Demographics. this poll has 75% White, 18% Black. In 2008 that was 70% and 20%. Also this poll has 22% olds and 12% youngs when in 2008 it was 11% olds and 22% youngs. It also has 53% female when it was 54% female. In other words it underweights women, blacks, hispanics and young people. If this poll were "unskewed" to match the 2008 demograhics it would have a huge Obama lead, not that I'm saying it should I'm just making a point about this silly party ID obsession.

Party ID is possibly the greatest factor in determining how someone will vote.

[J. J.]

But party ID is not a demographic. Party ID is a person's choice, just like voting intentions. Saying that democrats are oversampled in a poll is like saying the poll is wrong because it puts Obama ahead.

Party ID, which in some way is a measure of attitudes, is a demographic, i.e. "a single vital or social statistic of a human population."  http://dictionary.reference.com/browse/demographic


It obviously is a "social statistic," much like religious preference.

[J. J.]

Important point -

You must be logged in to read this quote.

Keep hope alive, fellas.

This is the most important point made by this poll. The job's report killed Romney's bounce!

They polled psychics?  The jobs report came out Friday, yet they said the numbers from Thursday were like those of Saturday.

[J. J.]

Then voting intention is a "social statistic" as well.

How they voted in the previous election might be, as would if they voted.  Except in cases where they have acted on intention, like requesting an absentee ballot, it would be the statistic the poll was trying to measure.

Are you seriously arguing that party ID is comparable to religious preference now?

Lief, do you honestly think that religious preference isn't a social statistic?