« Undecideds Souring on Bush | Main | General Election Cattle Call, September 16 »
Wednesday, September 15, 2004
General Election Cattle Call, September 15
Posted by Chris BowersNational Vote Projection
Bush: 49.01
Kerry: 48.99
Status: Toss-up
Polls Incluided: ABC, CBS, Fox, Gallup, ICR, Newsweek, Rasmussen and Zogby
Electoral Vote Projection
Bush: 284, 196 solid
Kerry: 254, 207 solid
States Changing Hands from 2000: NH to Kerry; WI to Bush
States Projected Under Three Points: FL, MO for Bush (38); ME CD-2, NH and PA for Kerry (26)
The last several weeks have brought me into closer contact and discussion with a wider range of statisticians, political scientists and pollsters than I ever imagined would happen for this election. The result has been a complete rethinking of the way I am going to project the Presidential Election. In an effort to improve the transparency, statistical validity, accuracy and purpose of the General Election Cattle Call, I have almost completely revamped my methodology.
The new results may not look possible based on current top-sheet polls results, but I feel pretty darn good about this system. And, yeah, I know it seems like I either slightly or massively alter the way I do these things ever other week, but I am always incorporating the new information I learn and discover about elections, demographics and statistics. Every alteration is a good one, I hope.
Right now, I am not projecting a large battleground. In fact, it is really just Pennsylvania and Florida. Kerry would need both in order to win.
Posted at 04:34 PM in General Election Cattle Call | Technorati
Comments
The weighting of all results by Party ID makes sense to me for statistical reasons, but I have no faith in the predicted turnout. What s the rationale for that split and how have predicted turnouts matched actually party turnouts in the past? Despite my concerns I still find the GECC the best predictor I've seen in that it properly reflects the federal system. This is why I am so concerned by this latest result.
Posted by: L_Dog at September 15, 2004 06:30 PM | Permalink | Edit Comment | Delete Comment
you can see historical breakdowns here:
http://www.hs.ttu.edu/hdfs3390/weighting.htm
Posted by: susan at September 15, 2004 08:34 PM | Permalink | Edit Comment | Delete Comment
A new poll done by CNN/Gallup has Kerry and Bush even in my home state of Minnesota. Of course another (I think more accurate) poll released by the Star Tribune today has Kerry with a 7 point lead. I personally think it's somewhere in the area of 4 points which isn't bad considering it was taken at the height of the Bush bounce.
Posted by: David Trinh at September 15, 2004 08:35 PM | Permalink | Edit Comment | Delete Comment
Bush is apparently trying to evade the second debate (town hall format, set in St. Louis). If he succeeds, a lot of Missourians will be very
insulted, making it a more likely pickup. And I still haven't seen a poll in Colorado with either a Bush lead or Bush with over 47%. With those, FL is unnecessary.
Posted by: science at September 15, 2004 09:21 PM | Permalink | Edit Comment | Delete Comment
The battleground is enormous. I think it's absolutely nuts to suggest that Florida and Pennsylvania are the only remaining states up for grabs. The post-GOP convention swing towards Bush showed me the electorate is not nearly as evenly polarized as what I thought it was a month ago..A continued Kerry implosion could make California and Illinois swing states. A bad September and October for Bush could put Virginia and Louisiana back in play. It's WAY to early to suggest the race is over in 48 American states. Everybody seems to put far too much stock in the poll of the hour. This thing is entirely up for grabs.
Posted by: Mark at September 15, 2004 09:24 PM | Permalink | Edit Comment | Delete Comment
Kerry would have to eat an baby on live TV for California or Illinois to be put in the swing state catergory.
Posted by: David Trinh at September 15, 2004 09:43 PM | Permalink | Edit Comment | Delete Comment
David T, the latest Survey USA poll shows Kerry up by only four points in Illinois. I suspect the sample isn't very representative, but it ais a warning sign that this electorate is far more fluid than what conventional wisdom a month ago suggested. I don't suspect Kerry will lose Illinois, but if Kerry enters October as badly as he entered September, many solid Kerry voters are likely to swing to Bush in the American tradition of not wanting to be on a losing team.
Posted by: Mark at September 15, 2004 10:39 PM | Permalink | Edit Comment | Delete Comment
I dont think Kerry really needs to get big numbers, i think he just needs to energize people to get out and VOTE. Like truman in '48, you dont need good poll numbers to win, you need a totally dedicated fan base to win. If bush has 60 percent support and kerry has 40 support, but half of Bush supporters actually vote and 80 percent of kerry supporters vote, kerry wins! (.6 X .5 = .3, .4 X .8 =.32) How can he do this?
One: Do well at the debates.
Two: Farenheit 9/11 on video
Three: rely on people like moveon.org. Because liberals hate bush SO much, he already has the edge here.
Four: The movie "Going Upriver" is going to probably piss some people off for continuing to talk about Vietnam, but i think its also going to energize democrats and possibly sway republicans. Find more about this movie here http://www.slate.com/id/2106758/ I actually think its a smart move becuase the people who have gotten pissed because of vietnam tactics have already switched sides (to witchever side they wanted) their card has been played (as has the swift vote people). Now its time (finally) for Kerry to play his card and answer these charges once and for all.
Can Kerry still lose? Yeah. This bush bounce proves Kerry doesnt have it wrapped up. But from now until election, Bush has nowhere to go but down, and the next 2 months are lined with stuff thats almost certain to boost kerry's numbers. Its probably going to be a close race, but i predict a kerry win.
PS if you are a democrat, go to www.johnkerry.com and do your part to help kerry get elected!
Posted by: Jared at September 16, 2004 02:09 AM | Permalink | Edit Comment | Delete Comment
What are your thoughts about Ohio? Do you
sense a Kerry win? Kerry should make a
bigger play for the Catholic vote, as it's
really up for grabs in the midwestern states.
Looking back on the 1992 election, the
Clinton strategists were masters. I recall
how Clinton delivered some foreign policy speeches at Georgetown University. Kerry, likewise, should give a speech at a Catholic university.
Posted by: Shar at September 16, 2004 02:52 AM | Permalink | Edit Comment | Delete Comment
I don't get it. Why does Kerry need to win Florida? If he wins PA, WI, ME, and either CO or AR or MO, then he wins. If we give credence to SUSA's claims in WI and PA, then shouldn't we also give it credence in MO and AR, where the same poll shows even races?
But Florida would make things easier.
Posted by: AWC at September 16, 2004 06:41 AM | Permalink | Edit Comment | Delete Comment
Yeah, I've read through the methodology, and I really don't get why Colorado isn't shown as a Kerry pickup or at least being within 3 points. The latest polls have all shown a dead heat or a 1 point Kerry lead.
I can only speculate that Chris is going on that historic party weighting formula. While that makes sense in a lot of states, I don't think it is a very good barometer in a state like Colorado where at least 30% of the population has moved there in the last ten years (and where traditionally a huge percentage of the population is moving in and out of the state all the time).
Posted by: Colorado Luis at September 16, 2004 12:05 PM | Permalink | Edit Comment | Delete Comment
OK, now I've read through it a third time, and it is the partisan weighting that has pushed Colorado into the Bush column in this projection. I don't think that's a valid criterion in the case of this state.
Posted by: Colorado Luis at September 16, 2004 12:09 PM | Permalink | Edit Comment | Delete Comment