Redstate

As much as I wish this was true, I think its a poor sampling.  650 people in CA is not nearly large enough.

In the end, if Bush can finish within 7 or 8 pts. of Kerry in CA, he can still win the popular vote.

I do know that here in NY, Kerry is polling between 53% and 59%, far below Gore's numbers in 2000...which means Kerry might be less likely to run up the score in the blue states.

By the way - is there anyway this DOESN'T get worse for the Dems after Arnold speaks in primetime at the convention?

Sure there is.  Arnold may be popular in California, but he's not a god.  Just how big do you think the pool of "undecided but heavily influenced by Schwarzenegger" is?

Isn't it reasonable to wonder if a mega-high profile appearence by the most popular (and moderate) republican in the state improves the Republican image in California?

If this poll is an accurate reflection of California voters, then this is very bad news for the Kerry campaign and very good news for the Bush campaign.  Simply put, the loss of California would be fatal to Kerry's ability to piece together a majority in the electoral college and thus his ability to win the Presidency.  If he were to lose California, he would have to win all of the states Gore won in 2000 and swing several key red states (Florida and Ohio) as well as convert a less electorally rich red state (Missouri, Virginia) to offset the loss of CA.  Granted, Bush will most likely not carry California, but the mere fact that this might be in play must create some panic in the Kerry camp.  At the very least, if this trend continues, Kerry will need to divert precious resources from the battleground states to solidify the most important state in his electoral strategy.  Doing so might cost him one or two key battleground states, which he cannot afford to lose.

As far as its accuracy, while I can accept the argument that the level of support for Bush might be too high, I don't believe it is too far off.  Recent polls (e.g. the CBS poll) suggest that the race is tightening, even moving in Bush's favor.  In reviewing the data in the poll, one of the things I find interesting is that Bush's support in CA hit its nadir at 41% back in early May, which is when, I believe, the Abu Ghraib prison abuse scandal surfaced.  Since then, his level of support has been steadily increasing in this poll.  I think that's a fairly accurate reflection of how the President is doing at a national level, too.  Since the spate of negative stories from Iraq has subsided, his numbers have steadily improved.  I expect this trend will be borne out in other statewide polls as well.  Barring any major unforeseen catastrophes or scandals, I believe Bush will begin to open a lead on Kerry since the support for Kerry, generally, is soft (see the high level of those who will "probably" vote for Kerry in this poll) and the Left has, to a certain degree, exhausted many of its ways to trash Bush.      

Sure, it's reasonable to wonder.  I wonder myself.  But that's not what you did.  You expressed a foregone conclusion that it will.

"Since the spate of negative stories from Iraq has subsided, his numbers have steadily improved.  I expect this trend will be borne out in other statewide polls as well.  Barring any major unforeseen catastrophes..."

I agree, but this is contingent upon us continuing to take it to the evildoers in Falluja, and doing what it takes to win back the Shia.  The worst news for Bush right now, I think, would not be a continued Iranian influence in Iraq, but an uptick in the number of Iraqis who want us out.  

And with Tehran threatening a preemptive strike against our troops, perhaps Tehran will provide the "October Surprise" that will push Bush over the top.

I hope the Bush campaign believes this poll and spends a lot of money and time in CA.

PLEASE!

I asked for a way in which Arnold doesn't help. Your response was "he doesn't".

I hope the Kerry campaign sluffs it off as you have and spends zero time and money anywhere in the state but Hollywood.

650 people is more than enough.  Standard state polls use that number.

If I were you, I wouldn't give alot of credence to a poll which shows Bush getting 28% of the African-American vote and more Republicans than Democrats voting in California.  I doubt very much if either campaign will take this poll seriously.

They oversampled Republicans, for one (45-36) and they're basing county/regions on this little a sample?

I hope its close to true...my heart says yes, but my body....oh my body..!

. . .if not your sentiments on this matter--given the small sample and its composition, it almost looks like someone is trying to convince GWB to waste limited funds and time in trying to carry the state, as he did in 2000.  If GWB wins CA, it will be in a LBJ/Reaganesque landslide--planning a pass play to leave the receiver in perfect position to spike the ball in front of the camera is not a strategy to emulate when the outcome of the game is in doubt.

For those folks who are dismissive of this poll, I suggest you visit the Survey USA election polls website.  I, too, was initially skeptical of these results, but after visiting their website to review the poll results for other elections and contests, I cannot see why this would be any less legitimate than polls conducted by other firms.  

 http://www.surveyusa.com/currentelectionpolls.html

... on US forces would be good for Bush?

just an honest question.

Survey USA does have a good track record, but that doesn't mean this poll is good.  Having significantly more Republicans than Democrats in their tabulation and 28% of African-Americans for Bush borders on self-parody by the polling firm.

If you look at the number of self-professed Dems and Republicans in the survey, it's near parity (240-230).  This was a random sample, and that's how the respondents identified themselves.  Thus, there are not significantly more Republicans than Dems in the survey as you state.  As far as Bush's level of black support, the self-identified black portion is 7% of poll respondents.  I don't know what portion of the California population identifies itself as African-American, but that seems within the ballpark.  28% of them stated they would vote for Bush.  Again, that's simply how the respondents replied.  Unless the methodology they used is flawed, I don't see how you can simply infer that these numbers are incorrect.

RealClear Politics tracks polls. They show only 3 for CA.  Survey USA and two others.  Latest for the two others show Kerry at +16 and +10

The composition of registered voters in the poll was 38%R and 37%D.  Voter registration in CA runs about 42%D and 35%R.  If the sample were deskewed, that would give Kerry about a 10 point lead, which is more in line with other polls.

Quick reasons why:

1. My gut feeling.  Not that I'm an expert, but it's a gut feeling.  Probably formed by the following in some small way.
   
2. Would legitimize more the threat posed by the Middle East as a whole and the need to have governments in that region that are not hostile to their people and to the U.S., which would sway US opinion that has been faltering on the wisdom of the Iraq war back in the direction of supporting it.
   
3. Polls have indicated without exception (that I've seen) that when the issue is US national security, Bush wins.  Big.
   
4. "Rally around the flag" all over again.  Just as no president has ever had as high marks as Bush did for a long time after 9/11 exactly because HE DID SOMETHING (in stark contrast to 8 years of doing nothing when attacked), if there was another major attack on us, I believe Americans would see Kerry as a return to Clinton's "tuck tail and run" strategy.  Kerry's own positions thus articulated that he plans on withdrawing from Iraq (much outlined here on Redstate) sure doesn't help that perception.

So that's a quick reaction to why i think Bush would benefit from an October Tehran Surprise.

650 for CA gives a margin of error of 3.5%. That's fine. However, remeber that by statistical definition 1 in 20 polls will fall outside that MOE. This is probably one of those ones.

RCP's poll average is just that -- an average of all the polls conducted.  It's not really sound because it does not account for the different methodologies each poll uses and lumps them together as if they are all the same.  

As far as the Survey USA results, to argue that its skewed is arguing its an unrepresentative poll based on the fact that it does not proportionally represent the level of party registration.  But, doing that would miss the point of a poll sample, which is to ask all registered/likely/potential voters who they would choose, not just the choice of those registered with a particular party.  If the point were to establish what we already know -- the level of support each party has by virtue of its respective registration levels -- then there would be no point to conducting a poll, since the purpose is to determine, at any given moment, the mood of those voting, regardless of regisration.  The fact that the level of disparity is not as large as you think it should be does not necessarily mean that its wrong.