22 October 2016

Statistikhengst's 2016 Electoral Landscape No. 1 (of 3): Clinton vs. Trump

Statistikhengst's Electoral Landscape No. 1:

Clinton vs. Trump
(Johnson, Stein, McMullin as sidenotes)

Data through October 21, 2016, 18 days before the election

Clinton 323 / Trump 215 / Margin: Clinton +108 EV
Clinton 323 / Trump 164 / Tossups 51 / Margin: Clinton +159 EV



This is my first electoral landscape prediction-output for the 2016 election. A lot of things have happened in my life, especially in 2016, and there was less time for detailed reports. Nonetheless, I have faithfully logged all of the polling data and also watched very carefully the events as they have passed. Making a report based on data ending at the end of the day 21 October 2016 is also beneficial as the Debate was Wednesday and no polling reflecting the final debate has come out yet - so this report will also be an excellent point of comparison to the next.

First off is to note that this election, in terms of statistics, is radically different than either 2012, 2008 or 2004, which were all very clearly 2-way races. This time, in 2016, whether you want to call it a protest vote or note, former GOP Gov. of New Mexico and current Libertarian presidential nominee Gary Johnson has often gone into double digits in polling, enough for major pollsters to justify 2 sets of statistics: 2-way and either 3-or-4-way statistics, the fourth initially being Green Party nominee Jill Stein. However, Ms. Stein was also polling between 2-3% in many 2012 polls and came in on election night 2012 with 0.36% of the vote - and for this reason, I see not much of a Jill Stein factor in this election at all. However, a fourth party (Independent) candidate named Evan McMullin, from Utah, may play a crucial role in at least one state. He is on the ballot in 11 states.

So, what does this mean for the calculations? It means that I needed to rethink how to calculate polling averages, because only taking values from a 2-way race in a year that at least has a feel of somewhat being a 3-way race would be wrong. Visa-versa also applies. Therefore I have taken the mean average of the 2-way and more-way stats to assign aggregate numbers to the states. So far, so good. In most states that post both sets of stats, there is not a huge divergence, but in a handful, the difference between the two is very large - and will be explained later.

First and foremost, here is all of the raw polling data to date, in excel-table form:

2016 Pres Polling - BATTLEGROUND/LANDSCAPE Data 21 October, 2016

The tables include hyperlinks to every single poll, the survey size, the MoE and the topline data, with the margins automatically calculated. The tables are in reverse chronological order, with the newest poll at the top and the oldest at the bottom.

The very first tab in the document lists the number of polls per state. Now, this calculation goes all the way back to February of 2013 (yes, I have been collecting them since way back then...), so not all of the Clinton or GOP figures mean a specific Clinton-vs-Trump matchup. However, I am a real believer in publishing the ENTIRE data-stream and not just cherry-picking.

In 2012, a number of "safe" states were never polled. This time around, due to 50-state polling from Morning Consult (R), Ipsos/Reuters, UPI/Cvoter and even Google (which I do not use, the survey sizes for many states are ridiculously small and I refuse to log half-assed polls with MoEs of +10!), every state plus DC has been polled a number of times.

Based on the polling results, which, as mentioned above, you can find all in the EXCEL DOCUMENT mentioned above and linked HERE again,  Based on the current aggregate date, I come up with the following electoral landscapes, one with a very tight definition of tossup, the other with a looser definition:







To-date, there are no special predictions for the congressional districts in either ME or NE, both of which do elector-splitting. There has been some competitive polling in ME-02 (Maine), but the survey groups have been very, very small. NE-02 (Nebraska) has not been polled at all, to my knowledge.

My methodology is similar to TPM's, but not identical. I take the average of all the last polls within 7-10 time, and with no repeater pollsters. Closer to election day, this time-frame will be reduced to one week, at least for the battleground states. Any average at +1% or less is pure tossup in the tightest model. Any average of +2% or less is the second, less restrictive model. Otherwise, I designate a prediction. I make no color distinction between states that are expected landslide states and states that are relatively close, for on election night, when the state is called, it is called, in one color, and that is that.

Either way, Democratic presidential nominee Hillary Clinton is way ahead, far much farther ahead than President Obama was at this juncture 4 years ago.


The distillation of the data from the EXCEL document is in two tables:

THE EVERYTHING TABLE:

Because it is so detailed, a legend:

The entire table is color coded: blue = DEMred = GOP.

Columns B and O: "Rank '12" is a listing of the partisan rankings for that state based on the results of the 2012 GE and is not a predictor of that state's performance in 2016. The Obama states go from 1-27 in bolded blue, the Romney states go from 24-1 in bolded red. Think of it as an hourglass, with the narrowest of margins in the middle.

Columns C and N are self-explanatory.

Column D: lists the current EV total for each state. If that total is underlined with another number in parentheses next to it, then this state has undergone a change in EV total since the 2010 Census. Those states are: NY, MA, MD, IL, WA, MI, NJ, NV, PA, IA, OH, FL, MO, GA, AZ, TX, LA and UT (18 states).

Columns E and M: the exact winning percentage margin from 2012, with a link in the title to Dave Leip's uselectionatlas.org website, as proof. Here the same LINK. His site is the DEFINITIVE repository of electoral data.

Column F: the current two-way polling aggregate for that state
Column G: the current 3 or 4-way polling aggregatge for that state.
Column H: the mean value of column F and G (if there are two values). If there are not two valuesk, then column F is repeated at column H, logically.

Column L is the end polling aggregate value from 2012.

Columns I and J are then automatically calculated:

Column I is the difference betwen the current polling aggregate (Column H) and the actual results from 2012 (Column E)

Column K is the difference betwen the current polling aggregate (Column H) and the end polling aggregate from 2012 (Column l)

Obviously for both of those columns, the larger the difference as a postive value, the better for the candidate.

So, here is the table:

Rank '12 State EV 2012 Margin '12 2016 2-way 2016-3/4-way 2016 PA aggregate Diff: '16 PA / '12
Diff: '16/'12 PA 2012 FINAL PA Margin '12 State Rank '08
1 DC 3 83,63 67,80
67,80 -15,83
-12,20 80,00 83,63 DC 1
2 HI 4 42,71 32,20
32,20 -10,51
5,20 27,00 42,71 HI 2
3 VT 3 35,60 30,30 28,00 29,15 -6,45
-6,70 37,00 35,60 VT 3
4 NY 29 28,13 23,60 24,00 23,80 -4,33
-1,60 25,20 28,13 NY 5
5 RI 4 27,46 19,80
19,80 -7,66
-1,20 21,00 27,46 RI 4
6 MD 10 26,08 21,80
21,80 -4,28
0,13 21,67 26,08 MD 7
7 MA 11 23,14 23,33 26,00 24,67 1,53
4,33 19,00 23,14 MA 6
8 CA 55 23,12 32,45 26,00 29,23 6,11
16,65 15,80 23,12 CA 10
9 DE 3 18,63 10,95
10,95 -7,68
-6,05 17,00 18,63 DE 9
10 NJ 14 17,68 15,57 14,00 14,78 -2,90
4,57 11,00 17,68 NJ 16
11 CT 7 17,33 15,00
15,00 -2,33
3,50 11,50 17,33 CT 11
12 IL 20 16,86 24,15 18,50 21,33 4,47
6,65 17,50 16,86 IL 8
13 ME 4 15,29 12,00 5,90 8,95 -6,34
0,97 11,03 15,29 ME 12
14 WA 12 14,78 10,75
10,75 -4,03
-2,99 13,74 14,78 WA 13
15 OR 7 12,09 9,30 9,00 9,15 -2,94
3,30 6,00 12,09 OR 15
16 NM 5 10,15 8,57 8,00 8,28 -1,87
0,57 8,00 10,15 NM 17
17 MI 16 9,48 9,35 9,50 9,43 -0,05
4,56 4,79 9,48 MI 14
18 MN 10 7,69 6,10 13,00 9,55 1,86
-0,15 6,25 7,69 MN 21
19 WI 10 6,94 6,65 6,66 6,66 -0,28
2,02 4,63 6,94 WI 18
20 NV 6 6,68 2,75 4,20 3,48 -3,20
-1,92 4,67 6,68 NV 19
21 IA 6 5,81 1,03 5,00 3,02 -2,79
-1,11 2,14 5,81 IA 23
22 NH 4 5,58 5,10 11,30 8,20 -0,48
2,60 2,50 5,58 NH 22
23 PA 20 5,38 6,32 5,93 6,12 0,74
2,12 4,20 5,38 PA 20
24 CO 9 5,37 7,65 6,25 6,95 1,58
5,78 1,87 5,37 CO 24
25 VA 13 3,87 9,00 11,75 10,38 6,51
8,23 0,77 3,87 VA 25
26 OH 18 2,97 0,56 1,80 1,18 -1,79
-2,68 3,24 2,97 OH 26
27 FL 29 0,88 3,10 2,50 2,80 1,92
2,49 0,61 0,88 FL 27

Total EV: 332
























NATL
538 3,85 6,73 6,75 6,74 2,89
5,51 1,22 3,85 USA --














Rank '12 State EV 2012 Margin '12


Diff: '16 PA / '12
Diff. 1 2012 FINAL PA Margin '12 State Rank '08
24 NC 15 2,04 3,58 3,33 3,46 1,42
1,18 2,40 2,04 NC 29
23 GA 16 7,81 2,90 1,43 2,16 -5,65
-5,35 8,25 7,81 GA 20
22 AZ 11 9,04 4,53 -1,02 (1,02) 1,76 -7,28
-3,14 7,67 9,04 AZ 18
21 MO 10 9,38 7,97 8,00 7,99 -1,39
-2,15 10,12 9,38 MO 22
20 IN 11 10,20 9,30 6,00 7,65 -2,55
-1,90 11,20 10,20 IN 28
19 SC 9 10,47 6,90
6,90 -3,57
0,90 6,00 10,47 SC 16
18 MS 6 11,50 10,00
10,00 -1,50
-5,00 15,00 11,50 MS 13
17 MT 3 13,65 16,95 10,00 13,48 -0,17
8,27 8,68 13,65 MT 21
16 AK 3 13,99 12,80 5,50 9,15 -4,84
-9,20 22,00 13,99 AK 6
15 TX 38 15,78 13,00 3,30 8,15 -7,63
-4,50 17,50 15,78 TX 15
14 LA 8 17,21 13,45 7,00 10,23 -6,98
-0,62 14,07 17,21 LA 8
13 SD 3 18,02 13,15
13,15 -4,87
1,15 12,00 18,02 SD 19
12 ND 3 19,62 14,65
14,65 -4,97
-2,85 17,50 19,62 ND 17
11 TN 11 20,40 14,35 10,00 12,18 -8,22
-3,65 18,00 20,40 TN 10
10 KS 6 21,72 14,00 11,00 12,50 -9,22
-5,80 19,80 21,72 KS 11
9 NE 5 21,78 17,25
17,25 -4,53
0,00 17,25 21,78 NE* 12
8 AL 9 22,19 17,68
17,68 -4,51
2,68 15,00 22,19 AL 5
7 KY 8 22,69 16,75
16,75 -5,94
2,75 14,00 22,69 KY 9
6 AR 6 23,69 17,68
17,68 -6,01
-9,32 27,00 23,69 AR 7
5 WV 5 26,75 17,35
17,35 -9,40
3,35 14,00 26,75 WV 14
4 ID 4 31,69 24,50
24,50 -7,19
-11,50 36,00 31,69 ID 4
3 OK 7 33,54 22,55
22,55 -10,99
-3,35 25,90 33,54 OK 2
2 WY 3 40,82 34,10
34,10 -6,72
-5,90 40,00 40,82 WY 1
1 UT 6 47,89 20,90 4,40 12, 9 / 4,4 #WERT!
-25,43 46,33 47,89 UT 3



You will see that some of the rows for the 2-way and 3/4-way polling aggregates are shaded in a bright orange-yellow color. Those are the states where there is a large disparity between the two-way and the more-way polling. Also, any value that is in ITALICS means that only one poll is involved, Any value that is not in italics means that more than one poll went into the average.

Be careful at jumping to too many conclusions when looked at current polling aggregates and the actual results. Example: HI. In 2012, the polling aggregate showed Obama +32.2, but he won by more than +42. Generally, very large landslide margins in polling can show a pretty big variance.

What we CAN see from the EVERYTHING TABLE is that of Obama's top 16 states, all of which went for him by double-digit margins, thirteen of them are also posting double digit margins for Hillary Clinton and the other three are also very close to the +10 mark. Similarly, of Romney's 20 states that went for him by double digits, 15 of them are posting double-digit margins for Trump and are every bit as safe for him as Clinton's safe states. I wrote 5 instead of 4 as it very much appears that the state of Utah is going to be a special case this year.

From the 16th to the 27th state from the Obama 2012 statistic, we see Clinton posting aggregate statistics right now pretty similar to either the Obama end-polling statistic or the actual result. For instance, in Michigan, which Obama won by +9.48, Clinton's current aggregate is: +9.43 (almost identical). In all-important Pennsylvania, Obama won by +5.38. Clinton's current aggregate is: +6.66 (end-timers, please do not freak out, it's just a number). Similarly, on the red side, some single digit states are posting similar values this time around, like MO: in 2012, Romney won MO by +9.38 and Trump is currently at +7.99, definitely in the ballpark. 

But based on the longstanding statistics now, it looks very much as if at least two states may definitely switch sides: IOWA, from DEM to GOP, and NORTH CAROLINA, from GOP to DEM. Now, OHIO just crossed over as a near tossup to Trump, but it has been in the Clinton column practically the entire time and may very well cross back over. And ARIZONA is also very close to the tossups.  To get a better view of all of this, here is the famous BATTLEGROUND TABLE, also in hourglass form:

A B C D F F G H I J K L M
State EV % of EC Polls 21.10.16 19.10.16 Shift 2012 2008 SWING (H-I) End polling 2012 Compare E – K Compare E – GE2012
UT 6 1,11% 2 of 6 12,90 5,07 7,83 47,89 28,02 19,87 46,33 -33,43 -34,99
SD 3 0,56% 2 of 2 10,55 10,55 0,00 18,02 8,41 9,61 12,00 -1,45 -7,47
MS 6 1,11% 3 of 3 9,90 9,90 0,00 11,50 13,17 -1,67 10,12 -0,22 -1,60
AK 3 0,56% 2 of 2 9,15 8,18 0,97 13,99 21,54 -7,55 22,00 -12,85 -4,84
TX 38 7.06% 2 of 6 8,15 8,25 -0,10 15,78 11,76 4,02 17,50 -9,35 -7,63
MO 10 1,86% 1 of 3 7,99 8,25 -0,26 9,38 0,13 9,25 10,12 -2,13 -1,39
IN 11 2,04% 2 of 3 7,65 7,50 0,15 10,20 -1,03 11,23 11,20 -3,55 -2,55
SC 9 1,67% 1 of 2 6,90 6,70 0,20 10,47 8,98 1,49 6,00 0,90 -3,57
UT*4w 6 1,11% 2 of 6 4,14 4,14 0,00 47,89 28,02 19,87 46,33 -42,19 -43,75
IA 6 1,12% 1 of 3 3,02 2,15 0,87 5,81 9,53 -3,72 2,14 5,16 8,83
GA 16 2,97% 4 of 6 2,16 1,88 0,28 7,81 5,20 2,61 8,25 -6,09 -5,65
AZ 11 2,04% 2 of 7 1,76 1,11 0,65 9,04 8,48 0,56 7,67 -5,91 -7,28
OH 18 3,35% 2 of 6 1,18 0,56 1,74 2,97 4,58 7,55 3,24 -2,06 -1,79
FL 29 5,39% 2 of 6 2,80 3,38 -0,58 0,88 2,81 -1,93 0,61 3,41 3,68
NC 15 2,79% 2 of 6 3,46 4,04 -0,58 2,40 0,33 2,73 2,40 5,86 5,86
NV 6 1,12% 1 of 6 3,48 2,98 0,50 6,68 12,49 -5,81 4,67 -1,19 -3,20
PA 20 3,72% 2 of 8 6,12 6,77 -0,65 5,38 10,31 -4,93 4,20 1,92 0,74
WI 10 1,86% 1 of 6 6,66 6,83 -0,17 6,94 13,90 -6,96 4,63 2,03 -0,28
USA 538 100,00% 7 of 21 6,74 6,64 0,10 3,86 7,26 -3,40 1,22 5,52 2,88
CO 9 1,67% 1 of 6 6,95 7,46 -0,51 5,37 8,95 -3,58 1,87 5,08 1,58
NH 4 0,74% 1 of 5 8,20 8,32 -0,12 5,58 9,61 -4,03 2,50 5,70 2,62
NM 5 0,93% 1 of 3 8,28 8,75 -0,47 10,15 15,13 25,28 8,00 0,28 -1,87
ME 4 0.74% 2 of 3 8,95 9,87 -0,92 15,29 17,32 -2,03 13,74 -4,79 -6,34
OR 7 1,30% 1 of 5 9,15 9,50 -0,35 12,09 16,35 -4,26 6,00 3,15 -2,94
MI 16 2,97% 1 of 4 9,43 12,25 -2,82 9,48 16,44 -6,96 4,79 4,64 -0,05
MN 10 1,86% 1 of 3 9,55 9,58 -0,03 7,69 10,24 -2,55 6,25 3,30 1,86
VA 13 2,42% 2 of 6 10,38 10,58 -0,20 3,87 6,30 -2,43 0,77 9,61 6,51


Currently, 8 states are in the battlegrounds, and the list may surprise you. Battleground is defined as an aggregate of +5 or less and tossup, if we want to go that far, is either +1 or less or maybe, to loosen things up, +2 or less.

First observation, and this is very different than 2012 or 2008: The national average Clinton vs. Trump is NOT in the battleground zone. This is a very important data point. With an average of Clinton +6.74 and a perfect spread between the two-way average (+6.73) and the more-way average (+6.75), Clinton is currently running about 7 points better than President Obama was on this day four years ago.

The eight official battlegrounds are:

Utah, Iowa, Georgia, Arizona, Ohio, Florida, North Carolina and Nevada.

Now, we are used to seeing Ohio, Nevada, Iowa and Florida in the battlegrounds. As of 2008, North Carolina also became a battleground and stayed that way in 2012 as well. But Georgia was Bush (43) +16 state in 2004. Arizona has only gone blue 1 time since 1968. And Utah is statistically the most conservative state in the USA. So, the list really is surprising.

Now, the states listed above (GOP) and below (DEM) the green battleground zone are either states that were once in the battlegrounds, were rumored to be possible battlegrounds or are statistically behaving differently than in 2012.  The most obvious example of a state that is no longer a battleground would be: VIRGINIA.

On this corresponding day in 2012 (18 days before the election), the polling aggregate in Virginia was: Romney +2.60 and on that day there were 11 battlegrounds, including the national average. On October 21, 2016, the polling aggregate for Virginia is: Clinton +10.38, 12.98 points better than Obama was doing exactly four years ago. It is quite obvious that Virginia's 13 electoral votes are not in the battlegrounds and even the Trump campaign knows this, for they have pulled out of Virginia.

But also, the Michigan statistic shows that Clinton is heading for landslide territory there as well - Michigan, a state that Donald Trump targeted as a way to get to 270 through the rust belt. Clinton's current Michigan statistic: +9.55. Obama's statistic from exactly four years ago: +6.68. Again, Clinton is outperforming where Obama was exactly at this juncture in time.

New Hampshire, the only state in the Northeast where Republicans have felt they had a chance of winning come electoral votes, is posting an aggregate of +8.20 for Clinton. Four years ago? It was a perfect numeric tie: 0.00. So, Clinton is running more than 8 points better now in NH than Obama did at this juncture in 2012.

In Pennsylvania, a state often very errantly called a "swing state" (it is not), the current statistic is: Clinton +6.12Four years agoObama +3.60. And the polling DNA of Pennsylvania is also very rich. There is, however, a disparity between the two-way and more-way values here.

Nevada is a state that has shown itself to be quite competitive this year and it is one of the few states where Clinton is not doing as well as Obama did four years ago. Atthis juncture (and based on quality, high-end polls), it's Clinton +3.48Four years ago? Obama +5.20.

Now, before I go to the surprise battlegrounds, let's look at the possible or likely "switchers" in this election.

Leading the pack as the state most likely to switch sides: NORTH CAROLINA, which is currently posting Clinton +3.46Four years agoRomney +2.00, so Clinton is running about 5.5 points better here than Obama did. Now, there's more to it than just that. North Carolina is to date the second most polled state in the nation, having received 88 polls and a total of 242 matchups, of which Clinton has won 148, and that in a state that George Bush (43) won by +13 in 2004. There is no doubt that there has been a techtonic shift in the demographics on the mid-atlantic states (Virginia, North Carolina, to a lesser degree, South Carolina), and those changes are making themselves noticeable in electoral politics. In 2008, Obama slipped ever so slightly ahead in the massive NC polling that year and his aggregate before election night was: Obama +0.3. He won on election night by +0.29. The end-polling aggregate in North Carolina showed Romney +2 and he won North Carolina by +2.04, so the polling aggregates out of the Tarheel State have been eerily of a bullseye quality; therefore,  I have not reason to doubt that Clinton has a lead here that Trump will not be able to overtake. She has been leading here for months on end. This is not a new phenomenon. When you combine Virgina and North Carolina's EV, you have 28 electors, just about the size of Florida.

On the other side, the state of IOWA is showing that it wants to go with Donald Trump. The current aggregate there is: Trump +3.02 (similar in strenght to Clinton's aggregate in North Carolina). Four years agoObama +2.76, so Trump is running almost 6 points better here than Romney was in 2012 and the demographics of that state tend to be a better fit for the tenor of his campaign: an extremely white state with lots of Evangelical voters, etc.

Between them is OHIO, my home state, a state that is the best bellwether of the nation. But the signs are that Ohio is showing resistance to the Clinton campaign and these figures are important. Right now, the aggregate in Ohio is: Trump +1.18 - which is knocking hard on the door to tossup territory, Four years agoObama +3.20, so Trump is running 4.5 points better in Ohio than Romney did at this juncture. But more importantly, in relation to the national numbers, Ohio is usually maybe 3 points under a Democratic winning margin. Right now, the Ohio aggregate is 8 full points to the Right of the national margin, a sign that the state may become less and less of a bellwether in the future. However, +1.18 is a margin that can shift again, very quickly. So, I would say, with respect to all sides that all bets are off when it comes to the Buckeye State. The problem for Trump is that Clinton, showing such strength in Virginia and North Carolina and Colorado - doesn't need Ohio to get to 270 - but Trump absolutely must win this state. BTW, CNN nailed the Ohio margin in polling in 2012, 2008, partially in 2004, in 2000, 1996 - so when CNN shows Trump ahead, I see a very strong possibility that he is ahead, at least for now.


Back to the Atlantic Seaboard: GEORGIA has sprung up as a real battleground and again, I believe this is mostly demographic shift and also a revitalization of the Democratic Party in the Peach State - a party that kind of went to sleep for about 2 decades. It's hard to believe that Jimmy Carter once won this, his home state, with over 60% of the vote.... and in 2004, George W. Bush (43) easily captured the state with over +16 in margin. Right now, in George, the aggregate is: Trump +2.16 but in more-way polling, it is only Trump +1.43, hard on the edge of the tossup zone. On this date in 2012, the aggregate was: Romney +8.00, so Trump is doing 5.8 to 6.5 points worse here than Romney - which tracks with Clinton being about 7 points ahead of Obama nationally, when you compared the two time-frames.

Now, Gary Johnson is doing better in this state than in many others, still getting about 9% of the vote whereas in many other states, he is down to 4% or so. This is a problem for both sides as Georgia, like Lousiana, has the "50% or runoff" rule, meaning that if either Trump or Clinton do not get to 50% +1 vote on election night, an automatic runoff election will take place at the beginning of December.  With that in mind, I sense that Republicans in the state who really cannot stomach Trump but who do not want to go through the rigors of a presidential runoff election, may give him their vote - especially if they see a Clinton electoral landslide mounting up, anyway.  BTW, to my knowledge, we have not yet had a presidential runoff election in Georgia. Were it to happen, I do think this would be the very first time.

Off to the SW of the USA: ARIZONA.  Nowhere in the USA is Trump's anti-Latino rhetoric causing a backlash as strong as it is in both Arizona and Colorado. The current AZ aggregate: Trump +1.76, BUT there is an enormous and meaningful difference between the two way calculations (Trump +4.53) and the more-way calculations (Clinton +1.02). Perhaps the most critical piece of evidence in all of this is that Republican pollsters are also showing the race as tied or with Clinton slightly ahead. Bill Clinton did narrowly win Arizona in 1996, and polling also showed him slightly ahead, a sure sign that Sen. Bob Dole (R-KS) was heading for a major defeat on election day. But that was a anomaly in a state that is generally considered an R+10 state. In fact, Romney won the state by +9.04% in 2012. This places Trump's current standing in Arizona at somewhere between 7.3 and 10.1 points worse than Romney's performance. Also, four years ago, to this day, the AZ statistic was: Romney +3.5 (but only based on 2 polls, one of which was a Democratic outlier pollster called Rocky Mountain). And just for shits and grins, the pollster who came the closest in AZ in 2012: PPP (D). which showed Romney +10.  

The point, however, is that Arizona, in a normal universe, should not even be competive. But this year it is and Clinton has a real chance of picking it up for the Democrats. That makes for 11 EV that Donald Trump cannot afford to lose.

All that being said, surely the largest surprise of the year is UTAH. Utah, a state that has gone reliably Republican every cycle since 1952, save 1964, 15 of 16 cycles. Just as sure as the sun rises in the East and sets in the West, Utahns vote Republican. But this year it is different. Mormons and minorities in this state are so put off by Trump, they are giving a fourth party candidate, Evan McMullin, a good hard look.

First, to the numeric nitty gritty. On paper, Utah shows the largest disparity between a 2-way aggregate (Trump +20.9 - already one half of the margin a Republican usually enjoys here) and a 4-way aggregate (Trump +4.40), making for a composite margin of Trump +12.9. Now, in most states, +12.9 would be a nice landslide margin, like Missisippi or Michigan, but in Utah, that is less than one-third of the usual margin and there is no doubt that the 2-way calculations in this case are absolutely worthless. Not only that, for the first time in my life, a FOURTH-PARTY candidate has won a presidential state poll:




I thought it would do good to post a screenshot of all polling out of Utah, a state that is, under normal circumstances, not polled at all. The screenshot proves that Trump has had a Utah problem for a long, long time, that this is not a new phenomenon. What IS new is the addition of Evan McMullin, a Utahn and also a Mormon, as a viable fourth-party candidate. We can see from the polling how McMullin's numbers have shot up and Gary Johnson's have receded.

I see four possible scenarios, none of them particularly good for Donald Trump:

Scenario 1: Trump goes and campaigns heavily in Utah (he was advised to do this and has not, to-date) in order to rally enough of the base to win in a 3.5 or 4-way race. This would mean less time for him in North Carolina, Ohio, Iowa and so on. Comparing 6 EV to 18? I bet he will go for the 18....

Scenario 2: Mitt Romney, Utah's favorite son, who has already openly said he will not vote for Trump or Clinton, comes out in support of McMullin, which would probably move the needle another 5 points or so for the former CIA-man.

Scenario 3: Gary Johnson, sensing that he has no chance, pulls out, and his supporters move toward McMullin,.

Scenario 4 (and I find this interesting): realizing that it is still unlikely for her to pick-up this ruby-RED state, but seeing an opportunity to deprive Donald Trump of Utah's 6 EV, the Clinton campaign could quietly, via surrogates, encourge the Clinton supporters to switch their voting preference to McMullin, which could move him to about 18 points ahead of Trump. This is a scenario that really COULD happen. There is nothing illegal or immoral about it, but it could be an excellent chess-move on the part of the Clinton campaign and I bet that their strategists are discussing this possibility right now, this weekend. Were I a strategist on her team, I would be doing exactly that.

So, on the Atlantic Seaboard, 28 EV that were once reliably Republican are either battlegrounds or firmly in the Democratic column.

And in the SW of the country, in two states that no one every really thought could be competitive, 16 EV are up for grabs.   That makes for 44 EV that George W. Bush never had to worry about.

And finally, to the most EV-rich battleground of them all: FLORIDA. Ahh, Florida, with 29 electoral votes, an incredible state with a very diverse population, which is a problem for Donald Trump's ultranationalistic campaign. The current aggregate: Clinton +2.80. Four years ago, on this day? Romney +0.38, so Clinton is running 3 points better now than Obama did back then. Can Trump turn it around in Florida? Possibly, but polling is showing that Cuban-Latinos and Asians (yes, there are also Asians in Florida) have completely turned their back on the GOP and polling out of Miami-Dade County, the Democratic powerhouse of the state, shows Clinton running 5 points better than Obama's end-statistic. So, at the end of the day, in Florida, it will come down to ground game and GOTV. But the Clinton polling reminds me a lot of the Bush 2004 polling from this state, where it was becoming obvious that the President really did have the upper hand in the state. Plus, Clinton and Trump are both banking votes in early voting in Florida, and it started during "THE TAPE".

The big view is that the battlegrounds have shifted again in 2016. Notice that we are not talking about Colorado or New Mexico or Wisconsin or Minnesota or Michigan or even Pennsylvania - states that have often been hotly contested. Instead, we are talking about the mid-to-southern Atlantic seaboard, Ohio, Iowa, Arizona, and unbelievably, Utah.


Nationally:



The tracking list:

12 October 2016, Clinton +5.52, -0.21
13 October 2016, Clinton +6.61, +1.08*
14 October 2016, Clinton +6.75, +0.14
15 October 2016 - no battleground report posted
16 October 2016, Clinton +6,44, -0.31
17 October, 2016Clinton +7.35, +0.92*
18 October, 2016, Clinton +6.91, -0.45
19 October 2016, Clinton +6.64, -0.27
20 October 2016 - no battleground report posted
21 October 2016, Clinton +6.74, +0.10
*Due to rounding, the difference is off by 0.01

On this corresponding date in 2012, the national aggregate was: Obama +0.44, so Clinton is currently doing 6.30 points better right now than Obama fared 18 days before the election in that year.


And some end-statistics: I first started collecting 2016 presidential polls at the beginning of October 2013 and since then I have logged 4,233matchups from 2,111 individual polls. You can find the breakdown to this stuff at the EXCEL table, in the tab that says "poll totals".

For past reference, here is the BATTLEGROUND REPORT from 18 days before the GE 2012, from 10/19/2012.  On 10/16/2008, 19 days before the election, I published my POLL CONVERGENCE and the DISTILLATION therefore (my equivalent of an Electoral Landscape back then). And her was the battleground report from 10/17/2008, 18 days before the election... numbers, numbers, numbers...


-Stat




No comments:

Post a Comment

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