Posted on 09/23/2012 8:55:35 AM PDT by LS
Ok, enough with the polls, especially when we now have some actual numbers to evaluate.
(Thanks to Freeper LD_Sentinel for pointing this out):
In OH absentee ballots are listed at a variety of websites, mostly at the county board of election websites. For example, in OH they show substantial advantages in Hamilton County (obviously, Cincy) for Republicans (+5500) and Franklin County (Columbus, +4000), and even Dem-heavy Starke (+700).
These sites list indies.
So for all you number crunchers in VA, FL, IA, WI, MI, NV, CO, and NH here is the formula:
1) Find your heaviest Dem county(counties), your heaviest R county (counties), and your swing county (counties). Look at the absentee ballot requests, note the difference in D/R/I.
2) Use the following assumptions: no R will vote for Obama; no D will vote for Romney; Is will split (according to almost every poll) at least 10% in Romney's favor (actually, most have it at least 15%, but we're going to side with the doubters here).
Apply #2 Indie numbers to #1 difference.
Now, if you really want to see where we are, find similar #s for 2008, and note the decline in D and the rise in R/I.
If we have these numbers, there will be no guess work about who is polling what. We will know, I suggest, one month out who will win.
Interesting in the results. Thanks Larry
Here’s 2008 for reference, as published by GMU:
2008:
Champaign Cnty 3,666
Cuyahoga Cnty 252,629
Franklin Cnty 207,243
Gallia Cnty 2,168
Greene Cnty 5,736
Knox Cnty 7,336
Montgomery Cnty 50,577
Muskingum Cnty 6,629
Ross Cnty 8,086
Seneca Cnty 4,156
Summit Cnty 73,920
Tuscarawas Cnty 9,339
Union Cnty 3,324
Also, a couple of really important ones (Hamilton and Warren) are missing.
It’s from the GMU project here: http://elections.gmu.edu/early_vote_2008.html
Looks like it’s total early vote by those particular counties. No idea why they didn’t list the other counties.
Of my 12 state project, only 2 are completed. The links to Google Docs shows the data.
This is my methodology.
(1) I go through each County-by-County within the state (or in the Case of NH, I have to go by township/wards since its reported at that level)
(2) I get the number of Dem/Rep votes in 2004 and 2008, compare the increase/decrease in the voter turnout to make a projection of the Dem/Rep voter population in that County for 2012 (flawed I know, since constant growth or decline is not certain, but it gets me close)
(3) I compare Bush’s performance to McCain’s as well as performances in the 2008/12 primaries/caucuses and estimate what Romney would have to pull within the county to be competitive at the State Level. For NV and other states that had big 2010 contests, I also looked at the voting % of the GOP Senate and Gov candidates to get better estimates.
(4) I apply my percentages Romney/Obama needs in each county to my estimated voter population for that county.
(5) When all county votes are tallied statewide, I review the Voter Party (ID) and % Makeup of the 2004, 2006, 2008, and 2010 Elections, as well as the share of Dem/Rep/Ind that Bush, McCain, and Obama all received to determine an estimate of what Romney or Obama would receive in 2012.
(6) Highlighted Counties in Yellow are Counties each Party (Bush/McCain) and (Kerry/Obama) held in 2004/2008, respectively. Purple Counties are counties McCain lost that Bush carried (McCain won no Counties carried by Kerry in IA and NV).
Based on this, I make my projections for that state.
So far, I have only done IA and NV. I will probably do CO and FL next. For OH, I will say that currently, Romney is beating Obama outside Cuyahoga County by about 50,000 votes (compared to 371,000 for Bush and just 214 votes for McCain. To win OH, Romney must exceed Obama’s vote total by about 150,000-200,000 votes, depending on well he performs in places like Hamilton County.
Iowa:
https://docs.google.com/file/d/0BxTof708DeHWVGNpc3JuVXd2Vk0/edit
Right now, if the voting model is correct, Obama is up about 23,000 votes in IA or 50.72% to 49.28% for Romney or 777,109 votes to 754,911. There seems to be just one suburban area too much in IA for Romney to overcome, and Republicans have been leaving Western IA due to the economy. Comes down to places like Black Hawk, and the Eastern side, though I think Romney will fare better in Des Moines and the Western suburbs. The Party-ID total seems to be about even and will probably be something like R(34), D(33) I(33). It’s been over 24 years since the GOP Candidate won the Independent Vote in IA, so to counter, Romney must win a larger share of GOP than Obama does of Democrats. But right now, I have then about even, with Romney getting 92% of GOP and Obama 93% of the Dems. The Indie’s are about 52-48 split for Obama. While the final vote totals may differ due to voter turnout, I think these percentages will hold for now if nothing changes in the race. IA will be close, but it appears Obama still has a slight advantage.
Nevada:
https://docs.google.com/file/d/0BxTof708DeHWUGU1YmFSc2VuZ2s/edit
Nevada has fewer counties and is a lot easier to do. Additionally, I added in the results of Angle/Reid and Sandoval/Reid races to get additional analyses on the percentages.
Outside of Washoe/Clark counties, the GOP candidate is going to clean up. Unfortunately, there just aren’t that many people there to offset the larger areas. Additionally, you see another problem with the Republican Electorate in NV. Bush managed to carry the state because he outperformed on the Hispanic vote, something McCain never came close to equaling. Angle did only slightly better than McCain, while Sandoval far outpaced even Bush, probably helped by his Hispanic heritage.
When looking at the numbers, it becomes clear that Romney most likely isn’t going to reach Sandoval’s percentages, but he should be able to run several points higher than both McCain and Angle, probably close to Bush, but probably not exceeding Bush’s numbers. This is where a Rubio might have helped the ticket.
Additionally, in NV when you go through county by county, noticed the numbers Ron Paul got in the small counties. There may be a sizeable Mormon population in NV, but there’s also a Libertarian slant to it as well. What these Paul supporters do may decide how competitive the state is. In all honesty, Romney needs 80% or more outside of Clark and Washoe to have a shot in NV. If Gary Johnson cuts into this, even if it’s in the Rural counties with few voters, I don’t see how Romney competes here.
The demographics seem to favor Obama here. Right now, I have him ahead 52.74% to 47.26% or 579,379 votes to 519,254. Looking at Part ID Data from 2004, 2006, 2008, and 2010, we are probably looking at a make-up of R(33) D(35) (I) 32. Bush got 93% of the R, while McCain got 89-90%. I see Romney getting 92%. Kerry got 90% of the (D) and Obama 94%, though I think Obama will get only 92-93% this time around, but with the Party-ID advantage, that means Romney would have to win the Indy’s here. The GOP Candidate has averaged about 44-46% of the Indy vote in NV. I don’t see that changing, at least not for now. That gives Obama a 60,000 or 4-5% advantage here. Despite the closeness, this is a state I’m not sure Romney has a chance at.
Eventually when I get a chance, I’m going to do CO and FL next.
Thanks for the links, I had no idea that GMU published statistics for 2010 as well.
It is worrisome that IA has had such a big absentee ballot push from the Dems, but requests don’t mean returns.
IMO, it’s still a couple of weeks too early to really pick at the early vote numbers, but I’ll be following and helping out with this thread where possible.
Wow, thanks for the excellent analysis, parks.
We should start a Freeper election number crunchers club.
On Saturday, the spouse and I received an important notification from Jon Husted, who is the current Secretary of State, taking note Via the United States Postal Service that we had moved out of State and to cancel our voter registration.
The point is two fold here....
1) They are cleaning the voter rolls in Ohio via the Postal Service records
2) and I wonder if the Secretary of States elsewhere are doing the same.
Also---not that I think you don't know your stuff, cause you do--- but how in the heck do any of these pollsters nationally keep saying that Romney is winning 10-15% more indies than Obama when, as the FL poll today shows and as you show with good logic he is either not winning indies or winning far fewer?
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.