26 September 2012

Bonncaruso's RED-STATE REPORT for 2012

Many have complained that there are not as many polls in 2012 as there were in 2008. At the end of the day, the numbers won't be that far off from each other, but one thing is for certain: up till now, there has been a real dearth of polls in some of the deepest of deep "red" states.

I have created a table of the 22 states that John McCain won in 2008, in descending order of margin, plus IN, which Obama won in 2008, but is the most likely candidate to return to the GOP in 2012. That makes 23 red states in this table, an EXCEL table very similar to the one being used for the nightly BATTLEGROUND REPORTS, which started last night, for instance, the one for 09/25/2012.

There will be a legend under this table, but one of the things that makes it different from the battleground table is that I have a column not only for the polling averages from 2012, but also a column for polling averages from 2011, because there are a number of red states that have not been polled in 2012 for the 2012 GE, but WERE polled in 2011. Those numbers, of course, are now very very old, but even they have a statistical story to tell.

Also in this table - and not in the battleground table, is the 20 year margin average for all of these states, from 1988-2008 (6 cycles, averaged). More about that later.

Here is the table:


StateEV% of EC2012 Avg.2011 Avg.20 Year Avg.20082004SWINGEnd polling 20082012 vs. 2008 EP2012 vs. 20 yr.2012 vs. GE 2008
WY30,56%

25,5332,2439,797,5523,00-23,00-25,53-32,24
OK71,30%30,00
19,5731,2931,140,1530,15-0,1510,43-1,29
UT61,12%42,0032,0030,8928,0245,5417,5224,0018,0011,1113,98
ID40,74%

26,7025,3038,1212,8223,00-23,00-26,70-25,30
AL91,67%15,00
12,6421,5825,624,0419,33-4,332,36-6,58
AK30,56%

21,3421,5425,554,0114,58-14,58-21,34-21,54
AR61,12%21,00
2,4319,859,7610,099,1011,9018,571,15
LA81,49%
15,005,7318,6314,514,1210,43-10,43-5,73-18,63
KY81,49%14,00
10,3016,2219,863,6412,401,603,70-2,22
TN112,04%7,0011,007,3815,0614,270,7914,00-7,00-0,38-8,06
NE50,93%11,0013,0022,3314,9333,2218,2919,00-8,00-11,33-3,93
KS61,12%
9,0016,2814,9225,3810,4617,00-17,00-16,28-14,92
MS61,12%
18,0014,1113,1719,696,5210,50-10,50-14,11-13,17
WV50,93%14,0021,000,0413,0912,860,237,666,3414,040,91
TX387,06%13,008,0012,8311,7622,8611,1013,000,000,171,24
SC91,67%-3,006,0013,358,9817,088,1012,55-15,55-16,35-11,98
ND30,56%17,00
15,948,6527,3618,711,3418,341,068,35
AZ112,04%7,337,007,708,4810,471,994,432,90-0,37-1,15
SD30,56%15,206,0010,998,4121,4713,067,507,704,216,79
GA162,97%13,507,009,055,2016,6011,403,859,654,458,30
MT30,56%7,0010,009,032,3820,5018,121,645,36-2,034,62
MO101,86%9,103,000,300,137,207,070,468,649,408,97
IN112,04%6,00
11,191,0320,6821,711,184,8217,194,97
Total19135,50%----
----------
--



Legend:

-all state names are linked to the large bios I did for each, all here in the PH Library.

-the number of EV and their respective % of the EC is self-explanatory. With IN, they add up to 191. Without IN, they add up to 180 EV according to the 2012 distribution of electors. In 2008, those same 22 states were worth 174 electors, 1 of which went for Obama (NE-02), so the GOP has automatically picked up 6 EV without adding a single state. In other words, the CORE GOP states just upped their percentage of the EC.

-the values for the 2012 and 2011 averages are there. Obviously, the 2011 averages are set in stone, but the 2012 averages may very well change and I hope to God that some of these states will get fresh polls. Any value in italics means that the state was polled only once, so that is the one-time polling margin and not an average. 

All of the blank fields you see mean that that particular state has not been polled at all in 2012: 

WY, ID, AK, LA, KS, MS


Plus, the following red states have either been polled only once in 2012 or have only one poll in the "average":

OK, UT, AL, AR, KY, TN, NE, WV, SD

So, only 8 GOP states have more than one poll in their current average:

TX, ND, AZ, GA, MT, MO, IN

-the 20 year average is from 1988-2008, but there is a caveat here: the three-way races in both 1992 and 1996 have caused all of these margins to go down some. If you look at the margins for these states going back to 1952, you will see considerably higher margin values for many of these states, especially the "11ers", which are charted HERE.

- the next three columns are the actual results from 2008, 2004, the "Swing" (margin difference, 2008 over 2004) and the end polling average from 2008. 

-the last three columns are very interesting, and they are the ones that fascinate statisticians from both sides: a comparison of the current polling averages to the polling from 2008, the results of the 20 year average and finally, to the 2008 GE.

First, I want to make it clear that I would personally be delighted were a ton of "red-state" polling to come in, for very obvious reasons: a healthy gene pool of a great amount of polling material allows for a much more accurate average, plus poll internals can tell us alot about how different people think in different states.

There is no doubt in my mind that Mitt Romney is going to win at least 22 of these 23 states, and probably all 23 of them.

But there are some interesting things to note:

UTAH could very well set a superlative record in 2012. Being the first Mormon presidential candidate, it is very conceivable that Mitt Romney wins UT with the largest percentage margin in that state's history. Up to 2008, UT was the number one state in the Conservative Rankings for 8 cycles in a row (1976, 1980, 1984, 1988, 1992, 1996, 2000 and 2004). It "dropped" to third place in 2008, but I see the very strong possibility that UT will be Romney's strongest state of all in 2012. The only poll of UT in 2012 give him a +42 margin over President Obama. BTW, the Deseret Poll is the gold standard for that state. They have been putting out good, solid, reliable results for years.

Also, BTW, for the first time, UT has begun to release voter registration statistics by party affiliation. Up until 2012, this was not the case. Here are the stats:

State VRDEMGOPINDNO AFFTotalDEM %GOP %IND %NO AFF %MarginMar %
UT 2012-09-014131358639168987967304814534539,04%43,98%0,68%46,31%50781034,94%


Utah looks a hell of a lot like a mirror image of Rhode Island:

State VRDEMGOPINDNO AFFTotalDEM %GOP %IND %NO AFF %MarginMar %
RI 2012-083017647784993134310572364941,70%10,76%0,13%47,41%22391530,94%

In Utah, only 9% of voters are Democrats. In Rhode Island, only 11% of voters are Republicans. Notice that the "no affiliation" % between these two states is almost identical and so is the % margin between the dominant party and the second party. Wow. Really like mirror images.

In the partisan rankings, UTAH, IDAHO and WYOMING have shared places 1, 2 and 3 more often than not and all three states have identical voting records going all the way back to 1916 (excepting WY in 1948, just once...), so even without polling in ID and WY to date, it is a sure bet that those states will be massive landslides for Mitt Romney. The question is one of margin. Will those states snap back to the configuration from 2004, where Bush enjoyed crushing landslides very reminiscent of Reagan 1984, or will those states remain with depressed margins, ala 2008? For this very reason, I would love to see some good polling out of these states.

OKLAHOMA was the most stabile state of 2008. It "swung" all of 0.15% more to the Right over the results from 2004. It is as if this state has just stood still in time. McCain won OK with +31.29% in 2008, Bush won OK by +31.14% in 2004. There have been 2 "Sooner" polls of OK in 2012. The first one showed Romney at +35, the second one shows him at +30. I would not be surprised if Romney wins by +31.40% or so here. Think of a cookie cutter that always puts out the same perfect form, every time... OK is an absolute CORE GOP state, as reliable as reliable can be. 


The polling from ARKANSAS, NORTH DAKOTA, SOUTH DAKOTA, GEORGIA, MONTANA, MISSOURI and INDIANA all show stronger numbers for Romney in 2012 than McCain earned in 2008.

ARKANSAS was one of the five states in 2008 to actually swing toward the GOP in a year where the other 45 (plus DC) swung more to the Democratic Party, and AK's swing was the strongest of them all. You can see that in the swing column. The five states in the Swing column with RED BOLDED values are the ones that swung more Republican. There have been two polls in AK, from Hendrix College: the first one from months ago had Romney at +24 over Obama, this one has him at +21 over Obama. Both values are higher than John McCain's win in 2008 and also higher than the polling average from 2008. Notice the very small 20 year average: AK used to be a core Democratic state and Bill Clinton landslided here in both 1992 and 1996. It has really tilted hard to the Right, starting in 2000, with no end in sight.

Both DAKOTAS were +8 margin states for McCain in 2008. Both show Romney in double digits over Obama, who has demonstrably lost ground here. ND is currently at more than double the result from 2008, SD is slightly less than double. This is absolutely in line with the history of these two states, which have identical voting records going back to 1920, but generally with larger GOP margins in ND.

Also part of Big-Sky country is MONTANA, which showed itself as very competitive for Obama in 2008. Not so much in 2012. McCain won the state by a narrow +2.38% in 2008, but he exceeded the end polling. Romney's average is currently at +7, which makes the state very solid, much as Michigan, Minnesota and Pennsylvania are looking for the President.

GEORGIA is showing massive improvement for the GOP over 2008, and with a current +13.50% average, could come close to Bush's +16 win in 2004.

MISSOURI is teaching us all a thing or two about the reality of "electoral shift". This state used to be, alongside Ohio and Nevada, one of the top 3 bellwethers of the nation. In 2008, as McCain won this state by +0.12%, Missouri lost it's bellwether status. Now, polls have been up and down here, but of the 20 polls taken  of the "Show-Me" state, Romney has won 18, Obama narrowly won one (May) and one was a tie. Romney is going to win Missouri and if the current +9.10 average is a guide, by a decidedly larger margin than in 2008 AND 2004. More likely than not, MO will "swing" GOP in the stats. 

INDIANA was one of the really amazing stories of 2008: a Democratic candidate capturing this absolute GOP bastion. The last time a Democrat did this, in 1964, he had to win nationally by about +24. Obama did it with a +7.26% national win. His +1.03% win in IN caused the largest cross partisan shift of that year: +21.73% toward the Democratic Party in the Hoosier State over the results from 2004 - a quantity very likely to spring back to the GOP in this year. But not guaranteed. The current average here is Romney +6, and that is comprised of 2 polls showing both +6. That is a middle-single digit margin. If Obama suddenly improves in the polling nationally, IN could theoretically become a battleground again, but there are so such signs of this at this time.

There are some states where Romney's polling is UNDER 2008:

KENTUCKY, TENNESSEE, NEBRASKA, SOUTH CAROLINA and ARIZONA.

SOUTH CAROLINA may be a fluke as there has only been one poll in 2012 and it is way too old. Very possible that Romney wins SC with +10 or more.

The changing demographic in ARIZONA can explain a +7.33% margin average. Hispanic influx, McCain not on the ballot this time. Lots of possible reasons.

But KY, TN and NE are indeed interesting. At +14KENTUCKY is just slightly under 2008, but above the 20 year average. Most people don't know, but KY used to be a major massive battleground state. Go look at the 1952 and 1956 elections and see how close this state used to be.

TENNESSEE is the big surprise for me. There is multiple polling showing Romney at +7 over Obama. That is half of McCain's margin from 2008, and somewhat inexplicable, as the state has clearly moved more to the Right. TN, along with KY, MO, LA, WV and AR, belongs to the "Clinton 6" - the six states that went for Clinton, but neither for Gore, Kerry nor Obama. I am sure that Romney will win TN easily on election night, but I am beginning to think that this really could be a single digit margin.

The other surprise is NEBRASKA: at +11, NE is under it's performance from 2008. And since both NE and KANSAS had almost identical margins in 2008 (14.93%, 14.92%, respectively), it is logical to assume that, ala North and South Dakota, that Nebraska and Kansas will go for Romney with similar margins once again. In fact, historically, this has been so. So, if Romney is at around +11 in NE, then I suspect he is at +11 or so in KS. Notice also the 2011 polling for Kansas...

The last state, the big Tamale, is of course: TEXAS. With 38 electors, the second largest state in the Union, TX is absolutely essential to a winning GOP coalition. I would really like to see much more pollling of the Lone Star State. Of the 6 polls of TX in this year, the margins have been anywhere between +7 and +20. At +13, TX is slightly above McCain's win from 2008, and almost exactly on par with the state's 20 year average. If, however, polling has not correctly calculated the hispanic voting block in all of this - for there are many, many young Hispanic Texans who have come of voting age between 2009 and today, then this could end up being a single-digit margin on election night.

-------------------------------------------------------------------------

FACIT: absolutely at least 180 and -more than likely- 191 EV (with Indiana) are in Mitt Romney's pocket and he is not going to lose them. But we are seeing conflicting statistical signs as to whether the crushing margins that were an essential part of a GOP national popular vote win, to counterbalance crushing margins that will come in for Obama in CA, IL, NY, for instance, are going to happen or not. 

One thing is for certain: more than 5 states are going to "swing" GOP in their margins this year, more than in 2008

No comments:

Post a Comment

Constructive comments and critique are always welcome. Please keep it polite and respectful.