Wednesday, October 31, 2012

Latest Michigan Poll - Anybody's game

There's a new poll from Glengarriff Group for the Detroit News and WDIV. It has rather interesting crosstabs.

The polls has this:
Obama - 47.5
Romney - 45

What really caught my attention is this.

This is all about turnout," said pollster Richard Czuba. "If Democrats turn out their voters, they win. If Republicans turn out more, they win. It's a pretty simple equation, as difficult as campaigns are."Romney leads Oakland County by 8.9 percent and Macomb County by 16 percent. Obama is up by 15.7 percent in Wayne County. That number does not include the city of Detroit where Obama owns nearly all votes.

IF, and that's a BIG IF -  IF either the Macomb or Oakland crosstab is accurate, we're in good shape. If BOTH Macomb and Oakland are accurate, (Wayne won't be accurate then) blowout win for Romney. How often do I predict blowouts?  I don't. Personally I think Macomb and Oakland's crosstabs are a little too good to be true, but I don't think Non Detroit Wayne County is a 16 pt deficit either. When Bush won Macomb by 1.5% and lost Oakland by 0.5%, he lost non-Detroit Wayne County by about 10.5%. The crosstabs don't make sense if Obama is supposed to be up by 2-3% I know Wayne outside of Detroit is getting bluer, but Mitt's not underperforming Kerry Downriver if he's up in Macomb 16%. He'll do better than Bush in the GP's and that NW corner of Northville, Plymouth and maybe Canton if he's up 8-9% in Oakland. Those areas are all similar to Novi, less so Canton. I expect Romney to underperform Bush in the airport areas (government unions and demographics) and Redford (Demographics).

An Oakland lead of 8.9% or Macomb lead by 16% is very significant. Here's the 04 and 08 results.

2008 GOP Dem Total GOP% Dem% Difference Diff
Macomb 187663 223784 419216 44.77% 53.38% -36121 -8.62%
Oakland  276956 372566 659068 42.02% 56.53% -95610 -14.51%
Detroit 8888 325534 335635 2.65% 96.99% -316646 -94.34%
Wayne 210964 334551 554691 38.03% 60.31% -123587 -22.28%
State 2048639 2872579 5001766 40.96% 57.43% -823940 -16.47%

2004 GOP Dem Total GOP% Dem% Difference Diff
Macomb 202166 196160 402410 50.24% 48.75% 6006 1.49%
Oakland  316633 319387 641977 49.32% 49.75% -2754 -0.43%
Detroit 19343 305528 325961 5.93% 93.73% -286185 -87.80%
Wayne 238407 294519 538767 44.25% 54.67% -56112 -10.41%
State 2313736 2479166 4839227 47.81% 51.23% -165430 -3.42%

For argument purposes, I'll use the 04 statewide votespreads with 08 Detroit votespread. That means there will be about a 200,000 vote deficit that needs to be made up. 2008 is an aberration. McCain quit, in a wave year and KO'ed the ground game for the GOP while Obama's ground game was at its height. 2010 is an aberration as well. I'm using 08 Detroit numbers because of race.

16pts in Macomb would be about 57-41-2 for Romney. Incredible numbers. That would be about a 64,385 vote spread. It would wipe out about 30% of the estimate without anything else.

8.9% in Oakland would be about  53.9-45-1.1. That would be about a 57,136 vote spread. It would wipe out over 25% of the estimate without anything else.

Keep in mind counties aren't in vacuums. Our own Livingston County closely mirrors Western and Northern Oakland County. It is consistently about 13.5% more Republican than Oakland County as a whole.

Snyder - Oakland - 60.10, Livingston 75.05 +15
McCain - Oakland - 41.94, Livingston 55.64 +13.7
DeVos - Oakland - 44.62, Livingston 57.05 +12.5
Bush 04 - Oakland - 49.32, Livingston 62.79 +13.5
Posthumus - Oakland - 48.30, Livingston 61.86 +13.5
Bush 00 - Oakland - 48.10, Livingston 59.14 +11

That is remarkably consistent. If Oakland is polling at 53.9% Romney, expect Livingston to poll at about 67.4%. The target to win statewide in Livingston County is usually between 63-64%. The spread in Livingston would likely be about 33,000 votes instead of 25,000. That extra 8,000 helps. St Clair and Monroe Counties tend to go in the same direction as Macomb County. Downriver has some similarities to Southern Macomb County. They lean democrat, but by how much? If that Macomb number really is 16%, Obama can kiss at least the southern half of Downriver goodbye, and probably Allen Park and maybe even places like Garden City as well in Wayne County.

Do I think these crosstabs are right? I have a hard time seeing Macomb go +16% for anybody, and +9 in Oakland is tough to believe as well. So is Romney being closer to McCain numbers than Bush numbers in Non-Detroit Wayne County. No way. 10% loss in Wayne County? I can see that with an improvement in Grosse Pointe, Plymouth, Northville and maybe Livonia, along with decreases in Redford, Belleville, Van Buren Twp, and Westland. I can also see 3% wins in Macomb and Oakland more likely. 5% win in either o those places will be good and give us a chance. 9% and 16% will signal a likely blowout IMO.

I'll have to see it to believe the poll crosstabs, but I think it's plausible. We can only control what we can control - but we CAN control what we can control. This will be won or lost with the ground game.  Polls are at best a scientific educated guess. No more and no less. Take it for what it is and not what it isn't (set in stone). 

Republicans and conservatives need to GOTV, work the polls, and control the things we can control.

No comments: