home

Gallup Tracker: McCain By 5

By Big Tent Democrat, Section Elections 2008
Posted on Wed Sep 10, 2008 at 12:25:37 PM EST
Tags: (all tags)
Share This: Digg! StumbleUpon del.icio.us reddit reddit

For the fourth day in a row, the Gallup tracking poll has McCain up, now 48-43:

The Sept. 7-9 average -- spanning interviewing conducted Sunday, Monday, and Tuesday -- finds little substantive change in the shape of the race compared to earlier this week, although each of the two candidate's level of support has dropped a point compared to the Tuesday's reported average. Nine percent of registered voters say they are undecided, refuse to state their preference, say they will vote for neither McCain nor Obama, or indicate they are voting for another candidate.

Here is a strange demographic factoid from Gallup's poll - Obama leads among women 49-42 while trailing among men by 49-44. And yet Gallup has McCain up 5. In a word, that simply is incredible. This Gallup poll seems faulty to me.

By Big Tent Democrat, speaking for me only

< What the Palin Pick Says About John McCain | Obama Happy With Hillary's Campaigning >

  • The Online Magazine with Liberal coverage of crime-related political and injustice news

  • Contribute To TalkLeft


    • Display: Sort:
    Gallup Tracker: McCain By 5 | 150 comments (150 topical)

    Readers Must Login or Create an Account to Comment

    The difference is (5.00 / 3) (#8)
    by Steve M on Wed Sep 10, 2008 at 12:33:34 PM EST
    48-43 represents the three-day average from Sept. 7-9.

    The numbers you quoted for men and women represent the weekly aggregate from Sept. 1-7.

    Since McCain did not take a lead, according to Gallup, until the latter part of the Sept. 1-7 period, these numbers seem perfectly consistent to me.

    They do? (none / 0) (#12)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:37:29 PM EST
    Really?

    Parent
    I see your point (none / 0) (#89)
    by Prabhata on Wed Sep 10, 2008 at 01:17:11 PM EST
    It's wrong.  There are more women voters and McCain must carry more men to make up the women deficit.  I think it's a typo.  McCain has a much better percentage of men.

    Parent
    The problem with your theory is this (none / 0) (#26)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:41:36 PM EST
    McCain has led by 5 for 4 straight days.

    From September 3-5, 4-6, 5-7, 6-8.

    So it includes the days September 3-7.

    Your explanation does not wash to me.

    Parent

    Hm (5.00 / 1) (#43)
    by Steve M on Wed Sep 10, 2008 at 12:49:21 PM EST
    I don't think you're quite right on that.  From Sept. 3-5, Obama was ahead by 2 points.

    So Obama was ahead from Sept. 1-3, 2-4, 3-5.

    McCain was ahead from Sept. 4-6 and 5-7.

    That seems perfectly consistent with the gender numbers coming out basically even for the overall Sept. 1-7 period.

    Besides, you can see what I'm talking about just by clicking the two links in your post.  The first says, quite clearly, that the numbers are for a Sept. 7-9 average.  The second says that the gender numbers you quote are for Sept. 1-7.

    Even if the gender numbers from Sept. 1-7 debunk something, you obviously can't use them to debunk a Sept. 7-9 poll.  The gender preferences from Sept. 1-7 tell us nothing about whom men and women preferred on Sept. 8, they tell us nothing about whom men and women preferred on Sept. 9, and they don't tell us a heck of a lot about Sept. 7, either.

    Parent

    Excuse me (none / 0) (#66)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:59:21 PM EST
    The number are the same for the third straight day.

    What you argue makes no sense.

    To wit, in the September 5-7 period, almost half of the 9/1-9/7 period, Gallup had McCain up 5.

    So on 5-7, then McCain should have had good numbers such that a weekly result of 49-42 women for Obama and 49-44 men for McCain is not squarable, UNLESS Gallup is predicting a 52-48 Men to women split in turnout.

    Parent

    Sigh (5.00 / 1) (#77)
    by Steve M on Wed Sep 10, 2008 at 01:05:24 PM EST
    And in the September 1-3 period, also almost half of the 9/1-9/7 period, Gallup had Obama up 7.

    For the middle part of the week, September 3-5, Obama was up 2.

    So if Obama is up 7 for the first half of the week, and McCain is up 5 for the second half of the week, why can't the overall men/women numbers for the week come out with a slight edge to Obama?

    Parent

    I think I can explain (5.00 / 1) (#91)
    by Manuel on Wed Sep 10, 2008 at 01:18:21 PM EST
    If you click on this link this link with matching time perid for the demograohic breakdowns, you'll see that for Sept 1-7 Obama is ahead 47% to 45%.  I do not know if that 2% lead is what one would expect given the corresponding gender breakdown.

    Parent
    And if you throw in the 4-6 period (none / 0) (#72)
    by Big Tent Democrat on Wed Sep 10, 2008 at 01:03:20 PM EST
    where McCain was up 3, then it makes it even more difficult to decipher.

    Parent
    You're confused ... (none / 0) (#92)
    by Robot Porter on Wed Sep 10, 2008 at 01:18:50 PM EST
    each number represents a weekly aggregate of that data set.

    That's why the male votes in that list add up to 103%, and the female number adds up to 91%.

    The numbers are only meant to be looked at as aggregates of individual data points, and, thus, are not comparable.

    Parent

    well now for sure I am (none / 0) (#110)
    by Lil on Wed Sep 10, 2008 at 02:20:43 PM EST
    but can someone explain (I'm almost afraid to ask) whata 5 poit lead for McCain means in terms of the oft mentioned "bounce"? Does he still have a bounce? Is it going up or down? I know your talking about over/under sampling, but I'm still trying to figure out which way it may be going.

    Parent
    That's more than the bounce (none / 0) (#113)
    by Cream City on Wed Sep 10, 2008 at 02:30:32 PM EST
    from the convention per several predictions I've seen, calculating that it would put McCain 1.4 points ahead of Obama on up to about 5.  (I prefer the careful calculations of those doing the lower predictions.)  So the rest may be a Palin bounce.

    And the impact of both at once means, I think, that we still have to wait and see for a few more days.

    Parent

    In Gallup Daily Tracking ... (none / 0) (#128)
    by Robot Porter on Wed Sep 10, 2008 at 04:56:02 PM EST
    Obama entered the convention 1% ahead, he reached a high of 8% ahead. So that was a 7% bounce.

    McCain is still at his post convention high of 5% ahead.  That's a 12% bounce.

    Generally, a bounce of 6% or more is considered good.

    It's called a bounce because it goes up and then goes down.

    Parent

    also the oddness (none / 0) (#133)
    by dailygrind on Wed Sep 10, 2008 at 06:23:16 PM EST
    of this cycle's convention line up and vp announcement makes it hard to argue which parts the bounce and which is not. ie, we are arguing compared to obama's own bounce versus the numbers prior to the conventions??

    Parent
    And we have Howard Dean ... (none / 0) (#148)
    by Robot Porter on Wed Sep 10, 2008 at 09:42:44 PM EST
    to thank for that.

    Parent
    well actually that may turn out to be a good thing (none / 0) (#150)
    by dailygrind on Wed Sep 10, 2008 at 11:23:19 PM EST
    because of the truncated event- it meant that we stepped on teh GOP's ability to build a narrative going into their convention like they did in 2004.

    Parent
    Yeah ... (none / 0) (#154)
    by Robot Porter on Thu Sep 11, 2008 at 10:21:51 AM EST
    because that whole Sarah Palin narrative didn't work at all.

    ;)

    Parent

    Men/women (none / 0) (#141)
    by DaveOinSF on Wed Sep 10, 2008 at 07:51:02 PM EST
    On September 3 in the tracking poll, Obama was up by 7.  Thus, you can assign Obama to be up 7 on September 1, September 2 and September 3.

    On September 7 in the tracking poll, McCain was up by 5.  You can assign him to be up by 5 on September 5, September 6 and September 7.

    That leaves September 4.  Judging by the days around it, it could have been anywhere from McCain by 2 or Obama by 1.

    Totalling up the seven days, it can be an average of Obama anywhere from 0.6 to 1.

    If over that seven day span Obama led among women 49-42 and McCain led men 49-44, a 50/50 M/F breakdown would have resulted in an Obama lead of exactly 1.

    If there are rounding issues, and obama's lead with women is slightly under 7 (but above 6.5) and McCain's lead with men is slightly above 5 (but less than 5.5), you can break down the electorate 47/53 M/F and still have Obama's overall lead be less than 1.

    In any case, there are lots of rounding issues and assumptions here but the math isn't nearly as far off as you suggest in your main post.

    Parent

    Adually (none / 0) (#142)
    by Big Tent Democrat on Wed Sep 10, 2008 at 07:57:04 PM EST
    Your comment makes my case.

    Obama by 1 is what it represents but Gallup says McCain by 5. And that is assuming a 50/50 split.

    I rest my case on your comment.

    Parent

    No (none / 0) (#143)
    by DaveOinSF on Wed Sep 10, 2008 at 08:00:03 PM EST
    Gallup says McCain by 5 from Sep 7-9.

    My math says Obama by 1 from Sep 1-7

    Parent

    it's rolling average of three days (none / 0) (#144)
    by dailygrind on Wed Sep 10, 2008 at 08:28:18 PM EST
    the numbers you are providing make no sense on that level.

    Parent
    Yes I know (none / 0) (#151)
    by DaveOinSF on Wed Sep 10, 2008 at 11:33:11 PM EST
    We know that there is a rolling average of O+7 over Sep 1-3 and of M +5 over Sep 5-7.  We can back estimate somewhere around O +1 to M+2 for Sep 4.

    Add it up:
    Single day performances:
    Sep 1 = O+7
    Sep 2 = O+7
    Sep 3 = O+7 (note that this results in a Sep 1-3 rolling average of O+7
    Sep 4 = even (guess)
    Sep 5 = M+5
    Sep 6 = M+5
    Sep 7 = M+5 (thus, Sep 5-7 rolling average = M+5)

    Taken together Sep 1-7 = ~O+1

    The data in the diary suggests men M+5 and women O+7 over that period (Sep 1-7)

    Assuming 50/50 split, it's O+1 for Sep 1-7, equal to the Sep 1-7 estimate calculated by breaking down the rolling averages.

    If you allow for rounding and say men were actually ~M+5.4 and women were actually ~ O+6.7, you can weight the poll to have 53% women and STILL come up with O+1 over the period Sep 1-7.

    The male/female breakdown over the period Sep 1-7 is NOT inconsistent with the individual rolling averages over that same period and certainly not inconsistent with M+5 for the Sep 7-9 period.

    Note that Sep 7-9 period result was actually realeased on Sep 10, but is marked on the graph as Sep 9.

    Parent

    to be clear- obama shouldn't be up by 1 (none / 0) (#145)
    by dailygrind on Wed Sep 10, 2008 at 08:29:25 PM EST
    for 1-7 and down by 5 the number of days he has been down or tied.

    Parent
    To be clear (none / 0) (#152)
    by DaveOinSF on Wed Sep 10, 2008 at 11:34:57 PM EST
    Yes he can.

    Obama was up by 7 on Sep1-3.  He was down by 5 on Sep 5-7.  If they were even on Sep 4, which is a pretty good guess, then Obama IS up by 1 over the period Sep 1-7.

    Parent

    So they def polled less women than men... (5.00 / 1) (#15)
    by LatinoDC on Wed Sep 10, 2008 at 12:38:24 PM EST
    shouldn't they consider polling at least 50-50? especially because it is pretty clear more women vote for Obama and McCain, in some cases, does better among men...this might have been a mistake Gallup had been making since the beginning...

    They're aggregates ... (none / 0) (#102)
    by Robot Porter on Wed Sep 10, 2008 at 01:45:44 PM EST
    you cannot extrapolate to full population averages.

    It's just a way of looking at what Obama and McCain's gender support looked like over the week individually, not in relation to each other.

    And since this was a volatile week the numbers look funky, and they're not much help.

    But they don't indicate what BTD is suggesting.

    Parent

    The Politico story supports your (5.00 / 4) (#20)
    by BarnBabe on Wed Sep 10, 2008 at 12:39:30 PM EST
    original argument. Could Clinton have Palin-proffed Dems

    Instead of a horse race we would have had the sure thing speaking for me only.We might have been 15 pts up now. And every headline I see where Obama is making a dig is at Palin. He is running against Palin. And Biden is a small blip on the screen. McCain is running under the radar.  

    The (5.00 / 2) (#34)
    by Ga6thDem on Wed Sep 10, 2008 at 12:44:38 PM EST
    most interesting thing about that article is the Obama campaign doesn't think they have a problem with women voters. So, obviously they will not be changing their strategy in this regards. They are going to continue to expect women to vote for them just like they did this past summer.

    Parent
    I don't think he's under the radar, I think (none / 0) (#23)
    by tigercourse on Wed Sep 10, 2008 at 12:40:34 PM EST
    we're just not bothering to shoot at him. Which is a big mistake.

    Parent
    BTD, YOUR TIME HORIZON IS OFF (5.00 / 0) (#93)
    by Smooth Jazz on Wed Sep 10, 2008 at 01:20:07 PM EST
    <Here is a strange demographic factoid from Gallup's poll - Obama leads among women 49-42 while trailing among men by 49-44. And yet Gallup has McCain up 5. In a word, that simply is incredible. This Gallup poll seems faulty to me.>

    This tracking poll covers 9/7 - 9/9 but that weird demographic you cited (O leads among women) covers 9/1 - 9/1, a period when Ohad built a sizable lead in this poll. Check you dates for that demo stat again.

    link (none / 0) (#105)
    by AlSmith on Wed Sep 10, 2008 at 02:07:04 PM EST

    can you give us the page link to where that is explained?

    I browsed around but I agree with BTD that I couldnt figure out how it added up. I kind of assumed that was was raw and one was reweighted but they dont tell you N, so you cant determine that.

    Parent

    Rasmussen says that McCain has pulled (5.00 / 0) (#106)
    by tigercourse on Wed Sep 10, 2008 at 02:11:51 PM EST
    ahead of Obama in New Mexico. They don't seem to have the poll up yet, but they mention it at the bottom of the Senate poll.

    I think Obama will win New Mexico, but it doesn't bode well for Colorado.

    I would need to know if this is their LV or RV (none / 0) (#108)
    by dailygrind on Wed Sep 10, 2008 at 02:16:11 PM EST
    model.

    Parent
    Likely voters ... (none / 0) (#129)
    by Robot Porter on Wed Sep 10, 2008 at 05:01:02 PM EST
    and the MOE is 4%.

    Parent
    yeah border line not such a good poll (none / 0) (#131)
    by dailygrind on Wed Sep 10, 2008 at 06:20:54 PM EST
    its okay, but if it's with a lot of leaners i am not impressed. Will wait for additional polling especially given ramussen's house effect.

    Parent
    I still say that ... (none / 0) (#147)
    by Robot Porter on Wed Sep 10, 2008 at 09:09:44 PM EST
    New Mexico is the Hamlet of the electoral college.  It cannot make up its mind.

    I don't want to wait up into the wee small hours of the fifth waiting for NM to be called.

    Parent

    Isn't it a problem (5.00 / 1) (#114)
    by Alien Abductee on Wed Sep 10, 2008 at 02:41:12 PM EST
    if the Gallup tracking poll, USA Today, and CBS News polls have changed their sampling since July to include an equal number of Republican voters as Democrats despite there being 11 million more Dems registered than Repubs?

    Or is this article confusing oversampling to get more accurate data from a smaller sample with over-representing?

    Could GOP ID really have spiked like this since July?

    Seriously all of this discusssion because (5.00 / 1) (#116)
    by Matt in Chicago on Wed Sep 10, 2008 at 02:53:21 PM EST
    Obama is losing in the polls?  I can't recall, did we have this kind of thread to determine if he was really as far ahead as the polls said?

    Trying to rationalize why the polls are wrong, is not going to address the underlying problem that the polls show his support is slipping.

    If we assume that the polls were right when we liked the result, we can't now assume that they are wrong simply because we don't like the result!

    Uh-oh, this polls cements it (5.00 / 1) (#118)
    by Dadler on Wed Sep 10, 2008 at 02:56:29 PM EST
    Poll Finds World's Preference Is Clear: Obama for President.

    He's done.  Tell America Obama win's the world's vote and we'll just swing the other way to piss them off.

    And I almost entirely serious.

    Seems right. (2.00 / 0) (#1)
    by cosbo on Wed Sep 10, 2008 at 12:28:29 PM EST
    The momentum has shifted to the McCain camp for now.

    Nooo it does not (5.00 / 3) (#3)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:31:23 PM EST
    More women than men vote. And even if it was equalt, Obama's lead among women is larger than McCain;s lead among men.

    This Gallup poll is nonsensical.

    Parent

    Historically (5.00 / 3) (#6)
    by Ga6thDem on Wed Sep 10, 2008 at 12:33:08 PM EST
    more women than men vote but lots of them may be sitting this one out.

    Parent
    Yep. Wondering about the gender split (5.00 / 3) (#9)
    by Cream City on Wed Sep 10, 2008 at 12:36:00 PM EST
    in that 9% undecided, too.

    Parent
    There's a new post-convention (5.00 / 3) (#22)
    by tootired on Wed Sep 10, 2008 at 12:39:46 PM EST
    movement among women called "N.O.T. - Nothing on Top". If you have 2 X chromosomes, you can probably get away with calling it "going into this election topless". I don't know if the group is large enough to make a blip in the polls, but it just might be. It's women who won't vote for Obama, but aren't ready to support McCain either. They are actually working quite hard on the down ticket.

    Parent
    They can work their fingers to the bone (none / 0) (#35)
    by coigue on Wed Sep 10, 2008 at 12:44:55 PM EST
    against a President McCain it won't get them the policies they want.

    Pretty short-sighted, if you ask me.

    Parent

    No offense meant (5.00 / 2) (#49)
    by tootired on Wed Sep 10, 2008 at 12:52:40 PM EST
    but they aren't asking any of us who support one candidate or the other for our opinions. We are irrelevant as are both of the presidential candidates. I have bumped into some of them while working on a local campaign. They're not trying to convert anyone either. They've just channeled all of their energies into the local races. They seem to have a lot of experience working on campaigns, and the candidates are lucky to have them.

    Parent
    Yep. Like a lot of us last night (5.00 / 6) (#67)
    by Cream City on Wed Sep 10, 2008 at 01:00:09 PM EST
    at a victory party for two local candidates for state legislature in my district -- one a woman, one a guy who gets it.  A lot of longtime Dems there last night said that we're putting most of our time and money into our local races.  And I especially am looking to anyone, like one of these candidates, who is a longtime worker for the party but a newcomer in terms of running for office, one who shares our view that the party needs to be rebuilt from bottom up.

    As for the top, I refused to mark my ballot yesterday for my super-delegate -- also a woman, and one for whom I worked hard, but who will not get my vote again.  Nor will Emily's List get my money again, as it helped her but did not hold her or others to promises they made to us.  

    Parent

    I am just saying that their downticket people (none / 0) (#65)
    by coigue on Wed Sep 10, 2008 at 12:58:32 PM EST
    and local community activists (like myself) will have an easier time if a Dem is in the WH.

    The last 8 years has felt like swimming upstream.

    Parent

    If you (3.00 / 0) (#42)
    by Ga6thDem on Wed Sep 10, 2008 at 12:48:20 PM EST
    have a candidate who is willing to concede every issue to the opposition you won't get what you want either.

    Parent
    opposition? (5.00 / 0) (#53)
    by coigue on Wed Sep 10, 2008 at 12:54:18 PM EST
    we're going to have the house and senate.

    I have been working at making changes at the local and state level these past LOOOONG 8 years. I am swimming upstream. I am sick of it.

    A Dem in the WH, any Dem, will make my work easier, better funded, and more effective. A Republican will make it harder.

    that's the bottom line.

    Parent

    So? (5.00 / 1) (#68)
    by Ga6thDem on Wed Sep 10, 2008 at 01:00:18 PM EST
    Obama is already conceding that McCain has the right ideas on the economy with his tax policy and I'm sure that the Dems would give him whatever Republican policy he decided to put together, right? Obama imo is too worried about being liked by the washington pundit crowd to do anything the least bit progressive. And the Dems in congress will likely go along with whatever he wants. Healthcare is already off the table. So nothing there.

    When Bill Clinton said that people are going to vote for a candidate who they agree with only 20% of the time but think he will get things done versus a candidate who they think will get nothing done who they agree with 100% he was right.

    I'm not advocating for McCain imo. I think his ideas are wrong for the country but I don't think Obama has a spine which could be even worse than McCain.

    Parent

    I disagree with everything you say here. (none / 0) (#71)
    by coigue on Wed Sep 10, 2008 at 01:02:47 PM EST
    let's just leave it at that.

    Parent
    You do understand that ... (none / 0) (#76)
    by Robot Porter on Wed Sep 10, 2008 at 01:05:09 PM EST
    the opposition to progressive policies doesn't only come from the Republican party.  There are other powerful groups fighting.  And those groups don't need elections to get and maintain their power.

    Parent
    of course. but those other opposition forces (none / 0) (#80)
    by coigue on Wed Sep 10, 2008 at 01:08:55 PM EST
    have more allies with republicans, and more enemies with Dems.

    You should also understand that the support of progressive policies exists outside of the Democratic party and will also be exerting influence. Any president has a choice of who to be influenced by. It will be relatively easy, in this climate, for any Dem president to be influenced by the the progressives. (note the use of the word "relatively".

    Parent

    So what is the percentage of the male electorate (none / 0) (#51)
    by magster on Wed Sep 10, 2008 at 12:54:05 PM EST
    in this poll? 60%? (I'm bad at math).

    Parent
    I can't tell (none / 0) (#56)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:55:18 PM EST


    Parent
    That is right (none / 0) (#61)
    by befuddledvoter on Wed Sep 10, 2008 at 12:57:12 PM EST
    The only way these results make sense is if women were drastically underrepresented in the sample.

    Parent
    I looked for the numbers by gender (none / 0) (#70)
    by befuddledvoter on Wed Sep 10, 2008 at 01:00:49 PM EST
    There are none.  All they have is percentage of each gender voting for each candidate.  This is bogus.

    Parent
    Men like Republicans (none / 0) (#82)
    by Prabhata on Wed Sep 10, 2008 at 01:12:06 PM EST
    Democrats are called girlie men (Schwarzenegger).  It's a historic trend and McCain keeps it.

    Parent
    Apples/Oranges (none / 0) (#139)
    by DaveOinSF on Wed Sep 10, 2008 at 07:32:07 PM EST
    The 48/43 tracking poll result is for September 7-9 data.

    The breakdown by sex that you cite is for the period September 1-7.

    Parent

    Here is Rasmussen latest update (none / 0) (#140)
    by Bob K on Wed Sep 10, 2008 at 07:35:03 PM EST
    on how the states are shifting. Link
    They have Obama 193 vs McCain 189

    Parent
    Yeah, I think the Gallup poll is giving (none / 0) (#2)
    by tigercourse on Wed Sep 10, 2008 at 12:30:34 PM EST
    McCain a bunch too many points. It doesn't match pretty much any other one out there.

    Give n a normal gender breakdown (none / 0) (#4)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:31:55 PM EST
    Obama would be up 2 or 3 in this poll.

    Parent
    Not a normal election, especially re gender (5.00 / 7) (#16)
    by Cream City on Wed Sep 10, 2008 at 12:38:29 PM EST
    of course.  Never before has this longtime voter and diehard Dem even considered sitting it out, in sheer disgust at the behaviors exhibited toward my gender in this campaign.  

    Parent
    Exactly (5.00 / 5) (#19)
    by TeresaInSnow2 on Wed Sep 10, 2008 at 12:39:24 PM EST
    me too, Cream.

    Parent
    Also (5.00 / 0) (#25)
    by TeresaInSnow2 on Wed Sep 10, 2008 at 12:41:34 PM EST
    Steve's right.

    The gender breakdown is an average from Sept 1-7.

    The poll is a rolling average from Sept 7-9.  

    The two are apples to oranges.  

    Gallup has designed things so that you can't do a proper comparison of the daily tracker to the gender breakdown.

    Parent

    Steve's wrong (none / 0) (#31)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:43:33 PM EST
    This is the 4th straight day Gallup has had McCain up 5.

    Parent
    You don't know for sure that he's wrong (none / 0) (#37)
    by TeresaInSnow2 on Wed Sep 10, 2008 at 12:46:07 PM EST
    We don't know anything for sure by comparing

    Sept 1-7 to 7-9

    apples to oranges.  

    September 1-3 included Obama's convention bounce, which also skews the numbers.

    Parent

    Sure we do (none / 0) (#54)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:54:44 PM EST
    We know that on September 4, Gallup reported the results for September 1 through 3,  Gallup had Obama up 4.

    We know that for the dates 9/3-5, Gallup had Obama up 2.

    We know that for the dates 9/4-6, Gallup had McCain up 3.

    We know that for the dates 9/ 5-7, Gallup had McCain up 5.

    We know that for the dates 9/6-8, Gallup had McCain up 5.

    We know that for the dates 9/7-9, Gallup had McCain up 5.

    We know that

    Parent

    Let's assume (5.00 / 1) (#69)
    by Steve M on Wed Sep 10, 2008 at 01:00:33 PM EST
    that tomorrow, Gallup tells you that McCain is up 5 for the dates 9/8-10.

    Heck, let's assume that next month, Gallup tells you that McCain is up 5 for the dates 10/8-10.

    Will you still claim that it can't possibly be true, because the gender results from 9/1-7 show an even race?  Because that's pretty much what you're doing, by arguing that the gender results from 9/1-7 mean that McCain can't possibly be ahead from 9/7-9.

    Parent

    Nooo (none / 0) (#83)
    by Big Tent Democrat on Wed Sep 10, 2008 at 01:12:31 PM EST
    I am arguing that since Gallup showed McCain up on the 4-7 period, the 9/1-9/7 numbers are not credible

    Parent
    two different polls (none / 0) (#62)
    by Shahryar on Wed Sep 10, 2008 at 12:57:30 PM EST
    I believe Gallup does a weekly poll and they also do a rolling poll.

    Parent
    Nope (none / 0) (#84)
    by Big Tent Democrat on Wed Sep 10, 2008 at 01:13:05 PM EST
    The 9/1-9/7 results are just a rollup of the entire daily polling.

    Parent
    Agreed (5.00 / 2) (#29)
    by Jellabean on Wed Sep 10, 2008 at 12:42:31 PM EST
    Not picking Hillary is going to come back to haunt Obama. Just think, Palin would not even have happened had Hillary been on the ticket. We'd be up against Mittens Romney or someone.

    Parent
    the only pundit who I've seen admitting to that (none / 0) (#39)
    by coigue on Wed Sep 10, 2008 at 12:46:27 PM EST
    is David Gergen.

    Parent
    Gergen always says (5.00 / 1) (#103)
    by oldpro on Wed Sep 10, 2008 at 01:56:14 PM EST
    exactly what he really thinks politically (within the bounds of good taste) and avoids personalizing issues.

    Last night he was just shaking his head at his co-panelists cheerleading responses to one question.  His was the only disagreement and he stated it firmly and twice.

    I'd go further but I'm having a senior moment and can't recall the exact question!  Oh well...

    Parent

    Dems who'll impede constitutional rights for all (5.00 / 5) (#101)
    by Ellie on Wed Sep 10, 2008 at 01:44:24 PM EST
    ... are a deal-breaker for this woman.

    That includes Dems who chase the hard right for months, knowing full well the decades-long agenda of persecution of the rhetorical menace du jour, and then turn around, in true fauxgressive fashion, to denounce that same hard right agenda when it looks like it might lose votes.

    Hated the sexist pile-on even more than the liberal bashing that Dems felt they've had to do for media for years, to be seen as Uniters Not Dividers. (I'll consider ALL my options this time, including whether breaking the glass ceiling so everyone can get the f*ck over it already is more important a core Dem value than having two lame Dems in the WH.)

    Indy by choice, since it's the only way my Ex(party) apparently gives a d@mn.

    Maybe others, with or without b00bies, feel the same way?

    Parent

    I second this statement (5.00 / 2) (#125)
    by Cream City on Wed Sep 10, 2008 at 04:18:03 PM EST
    in which you put the point so well:

    I'll consider ALL my options this time, including whether breaking the glass ceiling so everyone can get the f*ck over it already is more important a core Dem value than having two lame Dems in the WH.

    I'm reminded of the push-comes-to-shove point among abolitionists when deciding whether to abandon the Whigs or go to the options in the 1840s and then found a new party in the 1850s that won the WH in 1860.

    This time, we may be able to accomplish it without precipitating a civil war.  But war, it is.

    Parent

    Good point (none / 0) (#96)
    by Prabhata on Wed Sep 10, 2008 at 01:22:02 PM EST


    Parent
    Do you think (5.00 / 6) (#17)
    by TeresaInSnow2 on Wed Sep 10, 2008 at 12:38:41 PM EST
    the gender breakdown is going to be normal this time?

    I don't.

    Parent

    Maybe women are pissed off and tuned out. (none / 0) (#14)
    by cosbo on Wed Sep 10, 2008 at 12:38:20 PM EST


    Parent
    I tend to think women are very busy (5.00 / 5) (#85)
    by Militarytracy on Wed Sep 10, 2008 at 01:14:04 PM EST
    and intelligent enough to refocus on what Obama is going to do for them verses made up slurs if Obama will get off of his duff and deliver it up to them what he's going to do for them.  In the meantime most of us will continue to work a 40 hour week job and come home to put the bacon on the table, fold the towels, and make out the grocery list while falling asleep with pencil in hand.

    Parent
    True. (5.00 / 1) (#120)
    by cosbo on Wed Sep 10, 2008 at 03:07:25 PM EST
    I wish believed it was all women who was paying attention. There are women that I talk to today, that have no interest in politics are not registered to vote and have no interest in doing so. Sick right?

    As for what Obama can say that brings women to him and away from the first possible woman VP... I don't know. Really. Palin, like Obama, have a good narrative and on the surface appears to be your typical American woman/mother.

    George Bush won women last time despite the worsening global situation and the wars.

    Parent

    We didn't need all women before (5.00 / 2) (#126)
    by Cream City on Wed Sep 10, 2008 at 04:19:37 PM EST
    and we don't now.  We need enough women -- and, as before, a few good men . . . because it seems that a few of those are about all we ever get.  But if we have enough, that's all it took before, and that's all we need again.  We'll see.

    Parent
    Maybe Obama can cut into the white males (none / 0) (#41)
    by coigue on Wed Sep 10, 2008 at 12:47:08 PM EST
    with help from Scranton Joe.

    Parent
    Isn't it ironic (none / 0) (#104)
    by oldpro on Wed Sep 10, 2008 at 02:00:00 PM EST
    that the two male candidates have to rely on women to campaign for them to get the male vote?

    Parent
    Huh??? (none / 0) (#130)
    by coigue on Wed Sep 10, 2008 at 05:51:24 PM EST


    Parent
    Think about it. (none / 0) (#137)
    by oldpro on Wed Sep 10, 2008 at 07:04:46 PM EST


    Parent
    re read my comment (none / 0) (#149)
    by coigue on Wed Sep 10, 2008 at 09:48:29 PM EST
    scranton Joe is a man.

    Parent
    Yeesss, (none / 0) (#153)
    by oldpro on Thu Sep 11, 2008 at 12:31:42 AM EST
    but it's Hillary who is sent out to bring in the votes for Obama...and not just women but the lunchbucket brigade who are not voting for Obama so far.

    His gap is men...partiularly white men...who voted for Hillary in the primary...some were Dems, many were Independents and some were Republicans.

    No?

    Parent

    Faulty? (none / 0) (#5)
    by Ga6thDem on Wed Sep 10, 2008 at 12:32:00 PM EST
    Maybe but I remember picking over the internals of their polls in 2004 trying to discredit them but in the end they were right.

    So while a poll is still only a poll it tells us that McCain has gained ground and seems to be holding onto it.

    That is ridiculous (none / 0) (#10)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:36:09 PM EST
    The problem is clear as day.

    Gallup says Obama leads among women by 7 and trails among men by 5 and yet says McCain is up 5.

    That simply makes no sense.

    You must see that. Unless you don't want to see it, which is what I suspect.

    Parent

    Look (5.00 / 0) (#13)
    by Ga6thDem on Wed Sep 10, 2008 at 12:37:38 PM EST
    at Steve M's comment above. The gender numbers you are quoting are for a week while the top line numbers are from 3 days.

    Parent
    Steve's comment does not make sense (none / 0) (#30)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:42:55 PM EST


    Parent
    So is your contention (none / 0) (#40)
    by Jellabean on Wed Sep 10, 2008 at 12:47:05 PM EST
    Gallup is falsifying the poll for some arcane reason, or merely that you dont understand their methods?

    Parent
    No (none / 0) (#58)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:56:31 PM EST
    I think they are predicting much lower women turnout than history teaches us.

    Parent
    As a history teacher (5.00 / 2) (#74)
    by Cream City on Wed Sep 10, 2008 at 01:04:23 PM EST
    I applaud use of a useable past, as we say, when the past is a useful prologue.  I'm not sure that it is this time -- as we never have had an election like this time in terms of gender (as well as race).

    This is a reason why, fascinated by polls as I am, I think they all may be flying free of the usually useful predictors and models this time.  We will see.

    Parent

    Gallup doesn't weight by anything (5.00 / 1) (#78)
    by steviez314 on Wed Sep 10, 2008 at 01:05:42 PM EST
    so they're not predicting.  They are just basically making enough random phone calls to households a night to get 1000 registered voters.

    Over time, those random calls should get the right breakdown of party, gender, etc., but not necessarily over shorter periods of time.

    Whether men are more likley to answer the phone in a household before women, I have no clue.  I suspect in a household with parents and voting age child, the parent might answer first, skewing the sample older, but again I don't know.

    Parent

    My experience with Gallup Daily tracking (5.00 / 1) (#98)
    by Nike on Wed Sep 10, 2008 at 01:25:29 PM EST
    (which is to say they called my house once) was that aggressive or gendered phone answering would not matter.

    They wanted the adult over 18 who had most recently had a birthday. (So, it was my birthday, and I took the poll.)

    The gender bias would not be on who answers the phone, but there could be gender/class/age differences in terms of who is at home TO answer the phone.

    Parent

    I find that very hard to believe. (none / 0) (#88)
    by LarryInNYC on Wed Sep 10, 2008 at 01:16:14 PM EST
    If they were going to publish results based on a radical change in their voter model I think they would inform the poll's users of that fact.


    Parent
    BTD take a look at the White voter breakdown (none / 0) (#97)
    by Prabhata on Wed Sep 10, 2008 at 01:25:21 PM EST
    McCain carries both men and women by huge percentages.  Those numbers are diluted by the non-white.  There may be errors, but I think the overall numbers look right.

    Parent
    Can someone explain (none / 0) (#11)
    by Socraticsilence on Wed Sep 10, 2008 at 12:37:29 PM EST
    How the gender splits are exactly the same for the Sept. 1-7 and the preceding period (Obama +5 with women; McCain +7 with Men) but the overall poll has flipped, have men suddenly become a majority in the country?

    I think (5.00 / 2) (#18)
    by Ga6thDem on Wed Sep 10, 2008 at 12:38:55 PM EST
    it has more to do with who is going to vote than the actual population. If lots of women are sitting this one out it could explain the discrepancy.

    Parent
    Later today, Ras will have a New Mexico poll. (none / 0) (#21)
    by tigercourse on Wed Sep 10, 2008 at 12:39:34 PM EST
    In late August they had Obama ahead by 4. I'll be interested to see where he is now. I'm pretty sure he will win New Mexico in the end but if he's down by any big degree, it will tell me something about where he could end up in Colorado. That last sentence made sense in my mind.

    You are assuming there are equal numbers (none / 0) (#27)
    by Miserere mei on Wed Sep 10, 2008 at 12:41:36 PM EST
    of men and women. If there are more of one than the other then yes even almost even splits of the two will result in a higher overall lead of one of the candidates.

    Example:
    If we split women 50-45 and there are 100 women you get 50 and I get 45.

    Now if we split men 50-45 and there are 200 men you get 90 and I get 100.

    You total is 135, and mine is 145.

    That is just a simple  math example not tied to the poll itself.

    I do not follow your argument (none / 0) (#28)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:42:24 PM EST


    Parent
    I think other poster explained it well upthread. (none / 0) (#45)
    by Miserere mei on Wed Sep 10, 2008 at 12:50:58 PM EST
    You are seeing only a 2% advantage with McCain with men than Obama has with women so you are basically saying how can 2% add up to a 5% lead.

    It's because more men were polled over the three days than women. As always everything eventually averages out and there will be periods where more women are polled than men and if the splits remain the same Obama would have a 5% or so lead. That is why it is best to view polls over a long period rather than day to day because polling samples can affect the results.

    It is the same thing as when more Democrats are polled than Republicans and vis versa.

    Parent

    Maybe women aren't answering the phone (4.66 / 3) (#50)
    by Cream City on Wed Sep 10, 2008 at 12:53:44 PM EST
    if it's a phone poll?  Sometimes, when we're really ticked off, we just don't take calls.:-)

    Parent
    It appears to be a sample (none / 0) (#99)
    by Prabhata on Wed Sep 10, 2008 at 01:28:18 PM EST
    problem with non-white. The gender of the non-white is male heavy.

    Parent
    Isn't it true that women (none / 0) (#33)
    by Jellabean on Wed Sep 10, 2008 at 12:44:16 PM EST
    are more likely to vote than men? I thought I had read that.

    Parent
    In every election in the 28 years (5.00 / 1) (#38)
    by Big Tent Democrat on Wed Sep 10, 2008 at 12:46:24 PM EST
    that has been so.

    Parent
    Women majority of voters since 1952 (none / 0) (#46)
    by Cream City on Wed Sep 10, 2008 at 12:52:01 PM EST
    per many sources in women's history and elsewhere that I've seen.  I think you may be speaking of the "gender gap" in terms of women's political party preference?

    Parent
    Yes (none / 0) (#63)
    by Miserere mei on Wed Sep 10, 2008 at 12:57:40 PM EST
    But who votes vs who is polled on a particular day are two different things.

    The pollsters just call 'x' number of people and report the results. They don't say call 45/55 men/women because they don't know who they are going to reach or who wants to participate in the poll. Besides they can't predict what the men/women turnout ratio will be on election day. The call until they reach their target sample number and then report the results. Eventually everything more or less averages out. More or less I say because nothing is perfect and that is why polls are often wrong on election day.

    Myself I think they are great for watching trends and in the end they normally predict the winner but not always by the actual margin of victory.


    Parent

    my own poll (none / 0) (#36)
    by Dadler on Wed Sep 10, 2008 at 12:45:25 PM EST
    as a talkleft visitor, do you have any racial or gender prejudices that will impact your vote in the current race?  

    anyone?

    No just preferences (5.00 / 3) (#44)
    by smott on Wed Sep 10, 2008 at 12:50:01 PM EST
    ...for non-sexist candidates

    Parent
    Yes (5.00 / 4) (#47)
    by CST on Wed Sep 10, 2008 at 12:52:27 PM EST
    I would like to see an African American or a woman in the white house.

    Frankly, I would have loved to see an African American AND a woman in the white house.


    Parent

    sigh. Me too. (none / 0) (#59)
    by coigue on Wed Sep 10, 2008 at 12:56:31 PM EST


    Parent
    I am mightily ticked off at how (5.00 / 4) (#48)
    by oculus on Wed Sep 10, 2008 at 12:52:36 PM EST
    Sarah Palin, the VP candidate, is being incessantly attacked by some here and some in the press.  Not ticked off enough to vote for McCain though.  

    Parent
    I haven't done much back reading yet (none / 0) (#55)
    by Militarytracy on Wed Sep 10, 2008 at 12:54:56 PM EST
    Doesn't sound encouraging when you say things like that.

    Parent
    I am ticked (none / 0) (#60)
    by smott on Wed Sep 10, 2008 at 12:56:45 PM EST
    off that it was a trap to attack Palin and Obama supporters walked right into it.

    McCain swept their legs out by calling them on race-baiting early on, and they haven't recovered.

    And now they're getting called again on sexism with the lipstick bit.

    Parent

    Yeah I wish I could vote (none / 0) (#90)
    by Jellabean on Wed Sep 10, 2008 at 01:17:28 PM EST
    for Sarah and not for McCain hah!

    Just theoretically, what would happen if GOD forbid, something happened and John McCain were unable to run on Nov 4? Would Palin automatically be the candidate?

    Parent

    no (none / 0) (#109)
    by AlSmith on Wed Sep 10, 2008 at 02:17:06 PM EST

    I think this came up if Obama was disqualified by one of those lawsuits.

    The VP does not automatically become the nominee. The chairman and party insiders chose a new Pres candidate for their ticket.

    I believe that it is a separate state issue if there is still time for that person to be on a ballot.

    Parent

    adding (none / 0) (#111)
    by AlSmith on Wed Sep 10, 2008 at 02:21:25 PM EST
    If McCain dies on the 3rd, then its safe to assume he would stay on the ballot and when the electoral college actual voted the winner, the VP would ascend to the presidency.

    Parent
    I do (5.00 / 1) (#64)
    by Militarytracy on Wed Sep 10, 2008 at 12:57:55 PM EST
    I don't much want to vote for candidates that have racial and gender preferences in how they treat the humanity they must share space with.

    Parent
    Why, of course not. (none / 0) (#57)
    by coigue on Wed Sep 10, 2008 at 12:55:47 PM EST


    Parent
    Yes. (none / 0) (#132)
    by Alexi on Wed Sep 10, 2008 at 06:21:30 PM EST
    I will be using my vote to make it clear that it is not acceptable to gender-bash a female candidate, let alone define the middle class as ignorant trailer trash hillbillies.

    I will be doing that in every election from here on, regardless of party, until that message is received.

    And even if that vote impacts my pocketbook, I will consider it money well spent.

    Parent

    Diageo Hotline tracker has it tied 45-45 (none / 0) (#52)
    by steviez314 on Wed Sep 10, 2008 at 12:54:13 PM EST
    So, in other words, who the hell knows what's going on.

    54 days can have lots of stuff:

    4 debates
    another OBL tape
    war with Russia
    Citibank going under
    Market crash

    All can have much more impact on the elction than every tracking poll squiggle.

    I drove to Atlanta (5.00 / 1) (#79)
    by Militarytracy on Wed Sep 10, 2008 at 01:07:53 PM EST
    listening to C-span and the House Budget Com yesterday.  The stuff of nightmares baby.

    Parent
    Doesn't this follow (none / 0) (#73)
    by JAB on Wed Sep 10, 2008 at 01:04:08 PM EST
    The Ras poll that showed something like a 20 point swing with women from Obama to McCain?

    what does anyone think about (none / 0) (#75)
    by Jellabean on Wed Sep 10, 2008 at 01:05:02 PM EST
    The Bradley effect? Did Obama or did he not do better than polls in the primaries? Or did he do worse? Do we have to handicap the polls 5% to make up for the Bradley effect, i.e. people saying they will vote for a person of color, but when it comes to pulling the lever, they don't?

    Mostly (none / 0) (#81)
    by Ga6thDem on Wed Sep 10, 2008 at 01:09:31 PM EST
    the polling was accurate iirc as far as the percentage of the vote that Obama got. Where the "bradley effect" showed up, if in fact it was there, seemed to be in the undecideds.

    So if Obama is continually polling in the mid forties it could mean that's the number he gets on election day with McCain picking up the majority of undecideds.

    Parent

    My thought might be ungenerous (none / 0) (#86)
    by Jellabean on Wed Sep 10, 2008 at 01:15:35 PM EST
    But it seemed to me that Democrats in the primaries would be less likely to lie than independents and Republicans in the general, so he might do worse due to the Bradley effect more so than in the primaries.

    Parent
    Assuming there even is a Bradley Effect, (none / 0) (#95)
    by ruffian on Wed Sep 10, 2008 at 01:21:47 PM EST
    which I don't, shouldn't it shrink according to the closeness of the race?  Who would lie to a pollster about their support for McCain when he is a qualified candidate and it is perceived to be a close race?

    Parent
    BTD, the INTERNALS are for a DIFFERENT PERIOD (none / 0) (#87)
    by BigB on Wed Sep 10, 2008 at 01:15:53 PM EST
    The Internals are for Sep 1 to Sep 7 period.

    The top line numbers are for Sep 7 to Sep 9.

    And the numbers for (none / 0) (#94)
    by Big Tent Democrat on Wed Sep 10, 2008 at 01:20:41 PM EST
    the 4-7 where McCain also showed a 5 point lead?

    Parent
    Poll (none / 0) (#100)
    by pooks1976 on Wed Sep 10, 2008 at 01:41:51 PM EST
    BTD,
    I think they are right.  I think Obama had a big lead with women from probably 9/1 that diminished, when he was still enjoying his convention bounce, giving him a big lead in the average for the week, but when 9/1, then 9/2, then 9/3 were dropped from the 3 day average, it allowed McCain to take the lead.  You really can't get any hard data between a 7 day average and multiple 3 day averages for different periods.  I doubt Gallup, who may like polls even more than you, if possible, made such an obvious blunder.  

    Parent
    That's the real (5.00 / 1) (#119)
    by frankly0 on Wed Sep 10, 2008 at 03:00:38 PM EST
    point -- Gallup is not going to make a basic blunder like that.

    I'd sooner suspect a typo or a transcription error than that they'd make such a basic goof.

    They can't possibly have their very sophisticated software operating with such a gross bug (and that's what it would be) embedded in it.

    Parent

    You are right (none / 0) (#107)
    by dailygrind on Wed Sep 10, 2008 at 02:12:18 PM EST
    But there are several factors going into the polling data, including the following issues:

    a) Party ID: http://courses.ttu.edu/hdfs3390-reifman/weighting.htm

    One of the reasons it's hard to know what if any bounce McCain has is that no one (Bowers, Nate, etc) knows what party id is right now. Many of the pollsters shifted their id in favor of the GOP in August. That's why you had the Huffingtonpost article. I am not saying they are wrong, but I don't find the arguments by Bowers and Silver convincing. The weighting really matters for determining who will show up in the fall. Without this information we can't know for certain. I am unsatified with the answers I have seen from people thus far. Most of it is dismissive of the question or as Armstrong does it- tries to configure along the lines of 2004 turn out. This is problematic given the already known shifts in the electorate from 2004. Not to disparage them, but I feel like because they are so numbers/polls centric they feel the need to make excuses for the pollsters. Without transparency it's hard to know whether any of the assumptions or numbers are correct. Without clarifity on this issue- its hard to believe its anything more than reading tea leaves.

    b) Gallup has a history of being an outlier. This problem has been known since 1996. Anyone curious , who isn't just spinning, should look up the subject matter.

    Fla Dem writes about that here:

    http://www.openleft.com/showDiary.do?diaryId=8054

    Their polls varied wildly in 2004- sometimes erroneously in favor of Kerry when they shouldn't and sometimes the same for Bush.

    c) Gallup itself admits today that its numbers maybe a reflection of the South because in their cross tabs they make this point.

    See this: http://dailykos.com/storyonly/2008/9/10/10169/7586/567/593363

    The key data is this:
              Obama    McCain
    East       54       37
    MidWest    44       45
    South      39       51
    North      48       43

    Population wise this means that with 80 million people the Southern number while impressive for McCain doesn't translate into more electoral votes.

    d) You may also be right about your view of their gender breakdown.

    e) Finally, I am tending to think it's a tie just because most of the polling data points that way when one looks at as many of the polls as possible. There is just too much manipulation of the data to look at anyone number. It's partly a gut thing.

    Population shifts as well as political shifts (5.00 / 1) (#115)
    by Cream City on Wed Sep 10, 2008 at 02:47:17 PM EST
    may make some states difficult to predict, too.

    For example, my state of Wisconsin just was informed that the Census finds an increase in population since 2000 of more than 300,000.  Wisconsin was the second-closest state in the country in the 2000 election.  It was the closest state in the 2004 election.  Who knows if that means that the newcomers are trending red . . . or were trending red but some now may be moving blue. . . .

    Late in the decennial Census cycle as 2008 is, the population shifts -- more discussed for states like Colorado, so I've read here -- may make differences in the Electoral College count that also won't be that evident in these national polls.

    We really need those new state-by-state polls to see.  At this point, the Electoral College projections based on state polls that sometimes go back months are so close that just a shift of New Hampshire could toss this into the Senate -- which would work for Dems.  But it's darned close for comfort if it shifts more than that. . . .

    Parent

    Agree (none / 0) (#127)
    by dailygrind on Wed Sep 10, 2008 at 04:25:16 PM EST
    Turn out models this year will be difficult. I asked this question about being able to poll where one has to assume that 2008 is like 2004, but what if 2008 isn't like 2004? In other words, in 2006 we had a wave year- how does polling handle 2008 if this is also a wave year if the comparator is 2004? The answer was that it was hard to do that, if not impossible- that is make accurate assessments. Likewise, it's difficult to determine outcomes due to the increase in voter registration.

    I am really starting to take all of this with a grain of salt. If Obama is up, that's cool, but I am not buying it. If he's down,that's not so cool, but I am not buying it.

    I feel like this cycle there are so many variables that are new territory- that the honest pollster will say so. For example- if Obama has a record turn out in Virginia of AAs with the demographic shifts, then the polls wouldn't catpure this - especially LV models because those pressume the same turn out as previous cycles.

    The example I give of how we all got it wrong was the Iowa primary- we couldn't have predicted this.

    By the same token- peo also got the NH numbers wrong.

    This is why I am not sure what will happen. I just use them to peg whether things are close. I think they are. That means this is a ground game. This cycle- the ground game favors Obama. He spent the summer organizing unlike Kerry. McCain didn't spend his time building a ground game- thus he feels more like Kerry to me in a close race. Whereas Obama feels more like Bush. Going into the election, if the numbers are close or Obama slightly ahead after the debate- I think it favors Obama.

    This is also another factor- I will wait till the debates. In previous cycles, it took weeks for the GOP convention bounce to decrease. Kerry was down by 8-15 points under Gallup in 2004. He didn't come to parity until the first debate.

    I think the first debate favors McCain because he will use his POW status. I believe the last one favors Obama because its domestic policy- but then that's also a crap shoot because its about expectations and post debate spin wars.

    Parent

    More interesting permutations (none / 0) (#134)
    by Cream City on Wed Sep 10, 2008 at 06:46:08 PM EST
    and ramifications -- thanks, I enjoyed your post.

    We agree, I think, that national polling may be useful at best for trends, then.  And that the trend is not as good now for Obama as it was before.  Duh.  In other words, the race is tightening as we get closer to the election but are still before the debates.  Double duh.:-)

    Parent

    I agree, but I would also contextualize this (none / 0) (#136)
    by dailygrind on Wed Sep 10, 2008 at 06:58:36 PM EST
    by saying to McCain experienced a bounce as expected after his party's covention last week. The thing is, and I agree with Daily Kos about this (they are the least chicken little and have tended to place things in perspective which they dont' always do) : Mccain experienced a bounce, but it is already cresesting less than a week after his party's convention. It crested even with the most optomistically conservative reading at 5 points in the nationals. Bush crested much higher and much longer in 2004. It's bad for Obama that this happened, but it's not good long term for McCain at all. This is a sign of weakness and short term strength. This is why I look forward to the debates. I believe people are waiting to decide decisively then as they did with the 1980 race. Right now they are teetering and/or not paying attention. For mccain he can save himself by keeping it close and winning ina  close race, and for obama he can end this once and for all.

    Parent
    If the progressives were better mathematicians (none / 0) (#117)
    by lousy1 on Wed Sep 10, 2008 at 02:54:26 PM EST
    ... they would be conservatives.

    PA Poll (none / 0) (#121)
    by mmc9431 on Wed Sep 10, 2008 at 03:12:46 PM EST
    What troubles me is that so many polls are showing a decline for Obama. States where we had a significant lead are now polling very close. This election was never supposed to be a nail biter.

    Political Wire got an advance look at a new Strategic Vision poll in Pennsylvania that shows Sen. Barack Obama edging Sen. John McCain, 47% to 45% with 3% for other candidates and 5% undecided.


    As of June, this was a nailbiter (5.00 / 1) (#135)
    by Cream City on Wed Sep 10, 2008 at 06:49:24 PM EST
    when the Dems pushed out the candidate who had been winning since February 19 and went with the one who won for the first six weeks but then couldn't win Ohio, Pennsylvania, etc., etc.

    At that point, anyone who thought this was going to be easy thinks that we were running against Bush.  We never were.  Any other Republican is change, too -- to a lot of Republicans, independents, etc.

    Parent

    I think you hit the nail one the head (5.00 / 2) (#138)
    by Matt in Chicago on Wed Sep 10, 2008 at 07:06:11 PM EST
    We are running like we are running against Bush.  Hatred of Bush is clouding people's thinking... as if saying "Bush's third term" is enough just justify our winning the election.

    In my opinion, this is the third time we have tried the "Vote for me because I am NOT George Bush" campaign strategy.  And unfortunately, it seems as if the results will be the same for the election.

    American's want to be inspired by a leader who has hope for America... not one who says America is broken and only I can fix it.  That messages works for the faithful, but it turns off a whole lot of people.

    Parent

    I agree about Bush (none / 0) (#146)
    by mmc9431 on Wed Sep 10, 2008 at 09:01:31 PM EST
    Who ever is elected will be a change in that respect. That's why I've always thought Obama should tie the problems he discusses with the fact that these things occurred under Republican leadership. I also wish he would offer up a few solutions when he mentions the problems. (but then I wish I'd win the lottery too).

    Parent
    well (none / 0) (#122)
    by connecticut yankee on Wed Sep 10, 2008 at 03:17:38 PM EST
    Obama has rarely been far ahead and McCain breaks the surface every few weeks or months.  Only post-convention did Obama really wallop him.

    Which explains why the McCain camp is swinging so hard right now. They are terrified of another slide.  Theyve gotten air before and been unable to stay on the surface.

    Naw (5.00 / 1) (#123)
    by Ga6thDem on Wed Sep 10, 2008 at 03:20:28 PM EST
    this is what the GOP always does. They start pouring it on after labor day. It has nothing to do with fear or happiness it's just their MO.

    Parent
    Its not the overall numbers that are wrong. (none / 0) (#124)
    by elrapido on Wed Sep 10, 2008 at 03:58:08 PM EST
    Here is a strange demographic factoid from Gallup's poll - Obama leads among women 49-42 while trailing among men by 49-44. And yet Gallup has McCain up 5. In a word, that simply is incredible. This Gallup poll seems faulty to me.

    The faulty part would appear the poll's numbers for women.

    Gallup Tracker: McCain By 5 | 150 comments (150 topical)

    Readers Must Login or Create an Account to Comment