Wednesday, September 23, 2020

The Electoral College Map (9/23/20)

Update for September 23.


Thursday brought 17 new polls from eight states. Primarily those came from the core six battlegrounds -- Arizona, Florida, Michigan, North Carolina, Pennsylvania and Wisconsin -- but another couple of Trump-leaning toss ups -- Georgia and Ohio -- were also represented. Nothing from the flood of new data in eight of the most targeted states did anything to change the map -- still Biden 335-203 in the electoral vote tally -- but there was some important shuffling in the order of states on the Electoral College Spectrum. There will be more of that in the discussion below the Spectrum, but for now, here is a look at the cache of surveys that greeted Thursday.


Polling Quick Hits:
Arizona
(Trump 49, Biden 48 via ABC/WaPo | Biden 47, Trump 46 via Ipsos)
[Current FHQ margin: Biden +3.49]
Much was made early on today about the Republican tilt to the sample in the ABC/WaPo survey out of Arizona. [The same was true in Florida.] There are two things one can count on one hand about recent polling in Arizona. First, Trump leads have been few and far between, and second, the president has been at or above 49 percent in the Grand Canyon state just four times (in all of calendar 2020). That is not to say this one is necessarily an outlier. It could be the start of a new trend in the Sun Belt. Or, it could be an outlier. The Ipsos poll was the first in the state this year and is marginally more consistent with the FHQ averages. Biden ran a bit behind his average share while Trump was a couple of points ahead of his. But Ipsos was more in the range of typical results than was ABC/WaPo.


Florida
(Biden 50, Trump 47 via St. Pete Polls | Trump 51, Biden 47 via ABC/WaPo | Biden 47, Trump 47 via Ipsos)
[Current FHQ margin: Biden +3.42]
It was the first time in the field for both ABC/WaPo and Ipsos in the Sunshine state as well. But there is an established series from St. Pete Polls. That one can quickly be dispensed with however. Since the firm last conducted a Florida poll two weeks ago, neither the candidates shares of support nor the margin has changed. The focus, then, should probably be on the other two polls. But if one read the Arizona synopsis above, then the message was likely delivered on these Florida surveys as well. The ABC/WaPo poll looks like an outlier pending other future polls and Ipsos also had Biden running a tad behind his average while Trump was a couple of points ahead of his. So yeah, pretty much the same as Arizona. Importantly, however, the margin in Florida continues to tick down toward a tie, but still favors Biden for now.


Georgia
(Trump 50, Biden 45)
[Current FHQ margin: Trump +0.43]
Looking at the update in Georgia from Monmouth there was not a whole lot of change since the college pollster was last in the field in the Peach state at the end of July. Trump gained a point and Biden lost one in the low turnout likely voter model in that time. Even using the high turnout assumptions, the president remained stable since July as Biden lost a point. Neither assumption would have fundamentally changed where Georgia is in the rank ordering of states. Had it not been for the change in Ohio, the Peach state would have continued to be the closest Trump state on his side of the partisan line. It would have remained there had FHQ used the high turnout version of the results, but with the low turnout assumption, Georgia yields the distinction of being the closest Trump state to Ohio.


North Carolina
(Trump 45, Biden 44)
[Current FHQ margin: Biden +1.47]
The latest update of North Carolina from Harper Polling_ finds President Trump easing back into the lead in the Tar Heel state, but by the slimmest of margins. Back in early August it was Biden who held that same narrow 45-44 lead in the Harper survey. And that pre-convention poll, seemingly post-surge for the former vice president, was the high point for him in the series. Earlier polls in May and before had the president up by greater margins (and outside of the margin of error in April). That early Trump advantage has long since dissipated and been replaced in the current period by surveys showing a close race in North Carolina. Not only that, but the range of leads for both candidates has shrunk as well, all the while continuing to give Biden the advantage. But the Tar Heel state is the closest of the states among the Biden coalition right now.


Baldwin-Wallace (Great Lakes Poll #3)

Michigan: Biden +8 (Biden +3, Trump +/-0 since March wave)
Ohio: Biden +1 (Biden +2, Trump -3)
Pennsylvania: Biden +2 (Biden +2, Trump -2)
Wisconsin: Biden +9 (Biden +5, Trump -4)

The last Great Lakes Poll from Baldwin-Wallace came all the way back in March. In the six months since then, all four states have moved in Biden's direction. All have swung at least three points toward the former vice president, but none more than Wisconsin which shifted nine points in that time. That is fairly indicative of the polling of late in the Badger state. While in some other states, most notably Florida, there has been some narrowing of the gap between the two major party candidates, Wisconsin has continued to advantage Biden. Those double digit leads of the surge period for Biden has disappeared, but there continues to be a steady stream of mid- to upper single digit advantages that come in day after day. In this series, it has been enough to push Wisconsin past Michigan, but the Badger state is the only one out of order compared to the Spectrum below.


Change Research (late September battleground wave)

Arizona: Biden +6 (Biden +/-0, Trump -2 since early September wave)
Florida: Biden +3 (Biden +/-0, Trump +/-0)
Michigan: Biden +8 (Biden +2, Trump +/-0)
North Carolina: Biden +2 (Biden -1, Trump -1)
Pennsylvania: Biden +4 (Biden -1, Trump -1)
Wisconsin: Biden +9 (Biden +1, Trump -2)

There was also another Change Research update from those core battleground states. As has been the case in this twice monthly updated series, there just was not that much change from the last polls in the sequence. Like the B-W updates above, it was again Wisconsin that was out of place, looking marginally less competitive than Michigan. But Arizona was also out of order compared to what is depicted on the Spectrum. The Grand Canyon state ended up on the Biden side of Pennsylvania, and it was that survey that was more divergent from the FHQ averages than was true of the survey in the Keystone state.



NOTE: A description of the methodology behind the graduated weighted average of 2020 state-level polling that FHQ uses for these projections can be found here.


The Electoral College Spectrum1
DC-3
MA-11
(14)2
CT-7
(162)
NH-4
(253)
AK-3
(125)
AL-9
(60)
HI-4
(18)
NJ-14
(176)
PA-203
(273 | 285)
SC-9
(122)
IN-11
(51)
CA-55
(73)
OR-7
(183)
NV-6
(279 | 265)
MO-10
(113)
UT-6
(40)
VT-3
(76)
NM-5
(188)
AZ-11
(290 | 259)
MT-3
(103)
KY-8
(34)
NY-29
(105)
CO-9
(197)
FL-29
ME CD2-1
(320 | 248)
KS-6
NE CD1-1
(100)
ID-4
(26)
WA-12
(117)
VA-13
(210)
NC-15
(335 | 218)
LA-8
(93)
ND-3
(22)
MD-10
ME CD1-1
(128)
ME-2
(212)
OH-18
(203)
MS-6
(85)
SD-3
(19)
IL-20
(148)
MN-10
(222)
GA-16
(185)
AR-6
(79)
OK-7
(16)
RI-4
(152)
MI-16
(238)
IA-6
(169)
NE-2
(73)
WV-5
(9)
DE-3
(155)
NE CD2-1
WI-10
(249)
TX-38
(163)
TN-11
(71)
WY-3
NE CD3-1
(4)
1 Follow the link for a detailed explanation on how to read the Electoral College Spectrum.

2 The numbers in the parentheses refer to the number of electoral votes a candidate would have if he or she won all the states ranked prior to that state. If, for example, Trump won all the states up to and including Pennsylvania (Biden's toss up states plus the Pennsylvania), he would have 285 electoral votes. Trump's numbers are only totaled through the states he would need in order to get to 270. In those cases, Biden's number is on the left and Trump's is on the right in bold italics.

3 Pennsylvania
 is the state where Biden crosses the 270 electoral vote threshold to win the presidential election, the tipping point state. The tipping point cell is shaded in yellow to denote that and the font color is adjusted to attempt to reflect the category in which the state is.

All the new data on the day did not alter the map or the Watch List, but again, it was the order on the Electoral College Spectrum that saw some change after this round of surveys was added. Importantly, the collision course Arizona and Florida have found themselves on of late saw them cross paths with the Grand Canyon state moving one cell further away from the partisan line. The president may not need Arizona if he can pick off a state like Pennsylvania currently on the other side of Arizona in the order, but Trump in most scenarios that hew closely to the established order of states, absolutely needs Florida. That Florida is inching closer to tied, then, is, well, a bit of sunshine for the president.

As mentioned above Ohio jumped both Iowa and Georgia to settle in next to the partisan line. All three states are knotted together favoring the president by less than half a point. Their order is less consequential than their relative proximity to one another. And the talk of Wisconsin's recent margins in the discussions of both series of polls above is also not without consequence. The Badger state pushed past underpolled New Hampshire in the order, moving out of that most competitive middle column to join Michigan. The average margin did not change that much in Wisconsin, but that move in the order speaks to how it has stayed stable in Biden's column as some other states have shown some signs of tightening.



Where things stood at FHQ on September 23 (or close to it) in...
2016
2012
2008



--
NOTE: Distinctions are made between states based on how much they favor one candidate or another. States with a margin greater than 10 percent between Biden and Trump are "Strong" states. Those with a margin of 5 to 10 percent "Lean" toward one of the two (presumptive) nominees. Finally, states with a spread in the graduated weighted averages of both the candidates' shares of polling support less than 5 percent are "Toss Up" states. The darker a state is shaded in any of the figures here, the more strongly it is aligned with one of the candidates. Not all states along or near the boundaries between categories are close to pushing over into a neighboring group. Those most likely to switch -- those within a percentage point of the various lines of demarcation -- are included on the Watch List below.

The Watch List1
State
Potential Switch
Arkansas
from Strong Trump
to Lean Trump
Georgia
from Toss Up Trump
to Toss Up Biden
Iowa
from Toss Up Trump
to Toss Up Biden
Louisiana
from Lean Trump
to Strong Trump
Mississippi
from Strong Trump
to Lean Trump
Nevada
from Toss Up Biden
to Lean Biden
Ohio
from Toss Up Trump
to Toss Up Biden
Pennsylvania
from Lean Biden
to Toss Up Biden
South Carolina
from Lean Trump
to Toss Up Trump
Texas
from Toss Up Trump
to Toss Up Biden
1 Graduated weighted average margin within a fraction of a point of changing categories.

--
Methodological Note: In past years, FHQ has tried some different ways of dealing with states with no polls or just one poll in the early rounds of these projections. It does help that the least polled states are often the least competitive. The only shortcoming is that those states may be a little off in the order in the Spectrum. In earlier cycles, a simple average of the state's three previous cycles has been used. But in 2016, FHQ strayed from that and constructed an average swing from 2012 to 2016 that was applied to states. That method, however, did little to prevent anomalies like the Kansas poll that had Clinton ahead from biasing the averages. In 2016, the early average swing in the aggregate was  too small to make much difference anyway. For 2020, FHQ has utilized an average swing among states that were around a little polled state in the rank ordering on election day in 2016. If there is just one poll in Delaware in 2020, for example, then maybe it is reasonable to account for what the comparatively greater amount of polling tells us about the changes in Connecticut, New Jersey and New Mexico. Or perhaps the polling in Iowa, Mississippi and South Carolina so far tells us a bit about what may be happening in Alaska where no public polling has been released. That will hopefully work a bit better than the overall average that may end up a bit more muted.


--
Related posts:
The Electoral College Map (9/22/20)

The Electoral College Map (9/21/20)

The Electoral College Map (9/20/20)


Follow FHQ on TwitterInstagram and Facebook or subscribe by Email.

Tuesday, September 22, 2020

The Electoral College Map (9/22/20)

Update for September 22.


There are now just six weeks until election day. And this particular Tuesday, six weeks out, came with 10 new polls from nine states. It was a cache of data that updated the state of play in a number of battlegrounds while also providing new information in a pair of sporadically polled, but comfortably blue states. Across the board, however, the picture of the race for the White House at this moment remained one of stability. Joe Biden continues to hold durable leads in the important states that will determine whether he or President Donald Trump get to 270 electoral votes.

While Florida has drawn noticeably closer in recent days, those former blue wall states -- Michigan, Pennsylvania and Wisconsin -- are still seemingly stubbornly out of the president's reach by five to seven points. For every poll in any of those three states that emerges showing the race around, say, three points, there is another one (or more) that has the battle in those states within that five to seven point range. The president does not need all three of those states, but he will need one and to win all the toss ups to get there.


Polling Quick Hits:
Georgia
(Trump 47, Biden 47)
[Current FHQ margin: Trump +0.23]
The University of Georgia updated its outlook on the race for the Peach state's 16 electoral votes for the first time since a survey just before Super Tuesday back in March. The 51-43 Trump advantage then has closed to a tie, a finding that is consistent with where Georgia stands in the averages here. And as FHQ has said throughout the summer, if the discussion on election day about which states are the most competitive revolves around Georgia and those immediately around it in the order on the Spectrum below, then Biden is likely to be in pretty good shape.


Iowa
(Trump 47, Biden 47)
[Current FHQ margin: Trump +0.44]
Similar to the UGA series of surveys above, the Selzer polls in Iowa started out 2020 showing Trump up by a margin comparable to the results on election day in 2016. However, over time that lead has shrunk. The ten point advantage the president had over Biden in March around Super Tuesday fell to just one point three months later. Now, three months after that June survey, the battle for those six electoral votes in the Hawkeye state is a dead heat. Iowa is still slightly tipped in Trump's direction based on the full world of polling in the state in calendar 2020, but this series is indicative the trajectory of change there and where things may have leveled off.


Michigan
(Biden 46, Trump 41 via Market Research Group | Biden 49, Trump 44 via Ipsos)
[Current FHQ margin: Biden +7.42]
Both MRG and Ipsos have been in the field in Michigan this year, but the trajectory of change reflected in the two series is different over time. Biden gained in both sets, but Trump held pat in the low 40s in the MRG polls while his support grew at almost twice the rate of Biden's in the Ipsos set. The change in the latter can be at least partially pinned on the switch from a registered voter sample in April to likely voter screen being used now. But again, the margin in Michigan, one of those blue wall states from 2016, has been consistently in the seven to eight point range most of the summer. Polls that show a five point margin will cut into that over time, but slowly.


North Carolina
(Biden 47, Trump 47)
[Current FHQ margin: Biden +1.51]
The Ipsos survey of the Tar Heel state is the third from a closely contested toss up today that shows a race knotted at 47. Biden has been camped out around 47 percent support in the FHQ averages for a while, but Trump has generally trailed that but a couple of points. North Carolina, like Florida, has narrowed some of late, but the rate of change is much less dramatic in the Tar Heel state. The tight range of results has something to do with that. Biden's leads tend to top out now at +3 and the president's peak in August and September has been +2. Gone are those mid-summer polls during Biden's polling surge where his advantages stretched occasionally into the mid- and upper single digits.


Pennsylvania
(Biden 49, Trump 46)
[Current FHQ margin: Biden +5.16]
Ipsos also conducted a survey in Pennsylvania, but unlike in North Carolina, this was the firm's second in the Keystone state. As with the series discussed above in Georgia and Iowa, the Ipsos series in Pennsylvania found a tighter race now than in April. But that six point Biden lead then in a registered voter survey was cut in half under a likely voter screen. That transition may explain the bulk of that change rather than any natural narrowing. One thing that can be said about Pennsylvania polling in September is that Joe Biden is at or over 50 precent much less often than he was in earlier periods. The former vice president does not tend to be far under that majority threshold, but that is also indicative of the subtle changes that may be taking place in the commonwealth and elsewhere.


South Carolina
(Trump 50, Biden 44)
[Current FHQ margin: Trump +5.98]
What can one say about the state of polling in the Palmetto state? The latest Morning Consult survey there is very much in line with where polling has been all summer there. In fact, the five point Trump edge in the last Morning Consult poll in early August has barely changed. South Carolina is closer than it was in 2016, but it is clearly polling at a Lean Trump state with a margin just inside the Lean/Toss Up line on the lower end of the Lean range.


Vermont
(Biden 56, Trump 32)
[Current FHQ margin: Biden +27.44]
Braun Research brings the first glimpse of the race in the Green Mountain state all year. No, Vermont was never at risk of jumping the partisan line to join the Trump coalition of state much less shift into a less dark blue category. But it is good to confirm that with some 2020 data. But as was the case with the first Oregon poll to be released recently, this first foray in Vermont does not appear all that different than how things looked there in November 2016. That has not been true in all of the Strong Biden states, but it has in a handful of underpolled states in that category.


Washington
(Biden 58, Trump 36)
[Current FHQ margin: Biden +24.46]
The case in point is probably Washington. Biden has over performed Clinton in this solidly blue state while Trump has lagged behind his showing from four years ago. The latest Strategies 360 survey is indicative of that on some level. It has Biden right on his established FHQ average share of support. Trump, on the other hand, is running a little ahead of his own average but still almost 25 points behind.


Wisconsin
(Biden 48, Trump 43)
[Current FHQ margin: Biden +6.34]
Finally, Ipsos also surveyed Wisconsin and ended up with results that differ from some of the other Ipsos polls out today. The switch from registered to likely voter samples in Michigan and Pennsylvania drove much of the narrowing both states. But the story is slightly different in the Wisconsin series. The April registered voter survey from the firm had Biden up only three with more undecided voters and more support for other candidates. Both trailed off significantly in Michigan and Pennsylvania, but in Wisconsin, the undecideds remained stable and the other support dried up and ostensibly shifted to the two major candidates, but more so to Biden. That is part of the reason why the former vice president's margin grew from April to now in that sample transition, bringing the Ipsos poll in line with where the FHQ average margin in the Badger state currently rests.


NOTE: A description of the methodology behind the graduated weighted average of 2020 state-level polling that FHQ uses for these projections can be found here.


The Electoral College Spectrum1
DC-3
MA-11
(14)2
CT-7
(162)
WI-10
(253)
AK-3
(125)
AL-9
(60)
HI-4
(18)
NJ-14
(176)
PA-203
(273 | 285)
SC-9
(122)
IN-11
(51)
CA-55
(73)
OR-7
(183)
NV-6
(279 | 265)
MO-10
(113)
UT-6
(40)
VT-3
(76)
NM-5
(188)
FL-29
(308 | 259)
MT-3
(103)
KY-8
(34)
NY-29
(105)
CO-9
(197)
AZ-11
ME CD2-1
(320 | 230)
KS-6
NE CD1-1
(100)
ID-4
(26)
WA-12
(117)
VA-13
(210)
NC-15
(335 | 218)
LA-8
(93)
ND-3
(22)
MD-10
ME CD1-1
(128)
ME-2
(212)
GA-16
(203)
MS-6
(85)
SD-3
(19)
IL-20
(148)
MN-10
(222)
IA-6
(187)
AR-6
(79)
OK-7
(16)
RI-4
(152)
MI-16
(238)
OH-18
(181)
NE-2
(73)
WV-5
(9)
DE-3
(155)
NE CD2-1
NH-4
(243)
TX-38
(163)
TN-11
(71)
WY-3
NE CD3-1
(4)
1 Follow the link for a detailed explanation on how to read the Electoral College Spectrum.

2 The numbers in the parentheses refer to the number of electoral votes a candidate would have if he or she won all the states ranked prior to that state. If, for example, Trump won all the states up to and including Pennsylvania (Biden's toss up states plus the Pennsylvania), he would have 285 electoral votes. Trump's numbers are only totaled through the states he would need in order to get to 270. In those cases, Biden's number is on the left and Trump's is on the right in bold italics.

3 Pennsylvania
 is the state where Biden crosses the 270 electoral vote threshold to win the presidential election, the tipping point state. The tipping point cell is shaded in yellow to denote that and the font color is adjusted to attempt to reflect the category in which the state is.

For a day with two polls, the expectations may be high for some changes. But there just were not any or very many really. The map and Watch List remained unchanged from a day ago, and there was but one subtle shift on the Electoral College Spectrum. Iowa and Ohio switched spots with the Hawkeye state moving one cell closer to the partisan line separating the Biden and Trump states. But those two -- Iowa and Ohio -- are essentially tied at the moment within a one hundredth of a point of each other in their margins.

Pennsylvania continues to hold down the distinction of being the tipping point state in the order. Trump still has the ground between the partisan line and the Keystone state to make up in order to get back within striking distance of 270.



Where things stood at FHQ on September 22 (or close to it) in...
2016
2012
2008



--
NOTE: Distinctions are made between states based on how much they favor one candidate or another. States with a margin greater than 10 percent between Biden and Trump are "Strong" states. Those with a margin of 5 to 10 percent "Lean" toward one of the two (presumptive) nominees. Finally, states with a spread in the graduated weighted averages of both the candidates' shares of polling support less than 5 percent are "Toss Up" states. The darker a state is shaded in any of the figures here, the more strongly it is aligned with one of the candidates. Not all states along or near the boundaries between categories are close to pushing over into a neighboring group. Those most likely to switch -- those within a percentage point of the various lines of demarcation -- are included on the Watch List below.

The Watch List1
State
Potential Switch
Arkansas
from Strong Trump
to Lean Trump
Georgia
from Toss Up Trump
to Toss Up Biden
Iowa
from Toss Up Trump
to Toss Up Biden
Louisiana
from Lean Trump
to Strong Trump
Mississippi
from Strong Trump
to Lean Trump
Nevada
from Toss Up Biden
to Lean Biden
Ohio
from Toss Up Trump
to Toss Up Biden
Pennsylvania
from Lean Biden
to Toss Up Biden
South Carolina
from Lean Trump
to Toss Up Trump
Texas
from Toss Up Trump
to Toss Up Biden
1 Graduated weighted average margin within a fraction of a point of changing categories.

--
Methodological Note: In past years, FHQ has tried some different ways of dealing with states with no polls or just one poll in the early rounds of these projections. It does help that the least polled states are often the least competitive. The only shortcoming is that those states may be a little off in the order in the Spectrum. In earlier cycles, a simple average of the state's three previous cycles has been used. But in 2016, FHQ strayed from that and constructed an average swing from 2012 to 2016 that was applied to states. That method, however, did little to prevent anomalies like the Kansas poll that had Clinton ahead from biasing the averages. In 2016, the early average swing in the aggregate was  too small to make much difference anyway. For 2020, FHQ has utilized an average swing among states that were around a little polled state in the rank ordering on election day in 2016. If there is just one poll in Delaware in 2020, for example, then maybe it is reasonable to account for what the comparatively greater amount of polling tells us about the changes in Connecticut, New Jersey and New Mexico. Or perhaps the polling in Iowa, Mississippi and South Carolina so far tells us a bit about what may be happening in Alaska where no public polling has been released. That will hopefully work a bit better than the overall average that may end up a bit more muted.


--
Related posts:
The Electoral College Map (9/21/20)

The Electoral College Map (9/20/20)

The Electoral College Map (9/19/20)


Follow FHQ on TwitterInstagram and Facebook or subscribe by Email.

Monday, September 21, 2020

The Electoral College Map (9/21/20)

Update for September 21.


Changes (September 21)
StateBeforeAfter
Louisiana
Strong Trump
Lean Trump
The new work week commenced with an interesting line up of polls from mainly southern states and another round of polling from out of Maine. On the whole, it was a group of first time poll and it was good news for Joe Biden. But the swing though the Deep South from Tyson Group came with a number of surveys that do not exactly jibe well with the existing polling in those states. The Alabama poll is certainly one at which to look, but it did not have the impact that the latest Louisiana survey had. While the Yellowhammer state remained a Strong Trump state despite an unusually tight margin, the Pelican state survey was enough to nudge Louisiana back down below the Strong/Lean line into the very upper reaches of the Lean Trump category.


Polling Quick Hits:
Alabama
(Trump 48, Biden 44)
[Current FHQ margin: Trump +13.34]
The Alabama poll in question came from the Tyson Group and had Trump just four points ahead in a state the president carried by nearly 30 points in 2016. There has been a swing toward Democrats in 2020 polling, but it has not been that large. Not by a long shot. This one is an outlier, plain and simple. No Democrat has finished above 40 percent in any of the previous three cycles and no Republican has ended up below 60 precent. Now compare that the numbers above. Okay, moving on...


Florida
(Biden 46, Trump 44)
[Current FHQ margin: Biden +3.71]
The Tyson Group was also in the field last month in the Sunshine state with a survey that compared to the Alabama poll above at least ended in the range of both candidates' shares of support in other polls around the same time. It had Biden running a couple of points behind where he is now in the FHQ averages and Trump was within a point of his established average here. Overall, this one had minimal impact on where Florida is positioned. It remains on a collision course with Arizona, but with an average margin that remains just above that of the Grand Canyon state as of now.


Georgia
(Biden 49, Trump 46)
[Current FHQ margin: Trump +0.22]
The latest survey from GBAO_ comes form the Peach state and finds Biden narrowly ahead by three points. While that may be a bit more Biden-friendly than some other recent Georgia polling, it is not that far off. In fact, the difference is completely on the Biden number here. The firm nailed Trump's share established here in the FHQ averages, but had the former vice president three points out in front of his. Overall, this poll did not uproot Georgia from its position as the most competitive Trump toss up. It narrowed the president's advantage by a hair, but did not alter where the state is in the order depicted in the Electoral College Spectrum below.


Louisiana
(Trump 48, Biden 42)
[Current FHQ margin: Trump +9.63]
There has not been an extensive level of polling in Louisiana, but the Tyson Group survey of the Pelican state did not veer too far away from some of the work that has been in the field there in calendar 2020. But it would represent a 14 point swing toward the Democrats since 2016. Yes, that is about double the average swing at FHQ, but is not nearly as egregious as the Alabama outlier above. The rest of the Louisiana polling indicates a more modest swing. Yes, one that is above average at 10 points, but Biden's gains there are consistent with his average change across all states: around three points. Trump, on the other hand, has lost more than seven points in the Pelican state as compared to 2016, a shift that runs about three points over his average change across all states.


Maine
(Biden 51, Trump 39)
[Current FHQ margin: Biden +12.98]

Maine CD1
(Biden 55, Trump 34)
[Current FHQ margin: Biden +23.95]

Maine CD2
(Biden 47, Trump 45)
[Current FHQ margin: Biden +2.38]
The Suffolk of Maine just does not show that much divergence from either recent polling in the state (and in the two districts) nor the averages here at FHQ. The statewide poll and the data broken down by congressional district all paint pictures consistent with what has been established in the Pine Tree state. Needless to say, that does not really affect the average margins across any of the three jurisdictions there. Importantly, Biden remains narrowly ahead in the race for the one electoral vote in the competitive second district.


Mississippi
(Trump 50, Biden 40)
[Current FHQ margin: Trump +10.30]
What was said above about Maine can be extended to the Tyson Group survey of Mississippi as well. The only difference is that there have been fewer polls conducted in the Magnolia state. But this  poll like the Tyson poll of Florida was on par with other surveys that have been in the field in the state: Trump is ahead and comfortably so. And the findings are consistent with the candidates' average shares of support at FHQ. Little difference from the existing average margin means little change to the average. Mississippi remains just above the Strong/Lean line among the president's coalition of states.


North Carolina
(Biden 51, Trump 49)
[Current FHQ margin: Biden +1.54]
The one repeat survey of the day comes from Emerson out of the Tar Heel state. The college pollster was in the field with a survey a month ago and found just the opposite of what it finds in September. Rather than the president being up 51-49, it is the former vice president who holds the same advantage now. One could perhaps chalk that up to convention effects, but realistically, this is likely just noise in what has been a very close race for the 15 electoral votes at stake in North Carolina. But it is a very close race that has consistently given the edge -- but that very small one -- to Biden.


Texas
(Biden 48, Trump 44)
[Current FHQ margin: Trump +0.94]
Finally, it is not necessarily unusual to see Biden ahead in a poll of Texas, but the last Tyson Group survey has the former vice president at his ceiling so far in polling of the Lone Star state and Trump near his floor there. That means that both are in range of other surveys that have been conducted there, but the Biden lead in the poll is close to the maximum. That was enough to nudge the average margin in Texas below one point, but not enough to push the state over the partisan line. Like North Carolina above, Texas has been close but has been consistently tilted in Trump's direction.



NOTE: A description of the methodology behind the graduated weighted average of 2020 state-level polling that FHQ uses for these projections can be found here.


The Electoral College Spectrum1
DC-3
MA-11
(14)2
CT-7
(162)
WI-10
(253)
AK-3
(125)
AL-9
(60)
HI-4
(18)
NJ-14
(176)
PA-203
(273 | 285)
SC-9
(122)
IN-11
(51)
CA-55
(73)
OR-7
(183)
NV-6
(279 | 265)
MO-10
(113)
UT-6
(40)
VT-3
(76)
NM-5
(188)
FL-29
(308 | 259)
MT-3
(103)
KY-8
(34)
NY-29
(105)
CO-9
(197)
AZ-11
ME CD2-1
(320 | 230)
KS-6
NE CD1-1
(100)
ID-4
(26)
WA-12
(117)
VA-13
(210)
NC-15
(335 | 218)
LA-8
(93)
ND-3
(22)
MD-10
ME CD1-1
(128)
ME-2
(212)
GA-16
(203)
MS-6
(85)
SD-3
(19)
IL-20
(148)
MN-10
(222)
OH-18
(187)
AR-6
(79)
OK-7
(16)
RI-4
(152)
MI-16
(238)
IA-6
(169)
NE-2
(73)
WV-5
(9)
DE-3
(155)
NE CD2-1
NH-4
(243)
TX-38
(163)
TN-11
(71)
WY-3
NE CD3-1
(4)
1 Follow the link for a detailed explanation on how to read the Electoral College Spectrum.

2 The numbers in the parentheses refer to the number of electoral votes a candidate would have if he or she won all the states ranked prior to that state. If, for example, Trump won all the states up to and including Pennsylvania (Biden's toss up states plus the Pennsylvania), he would have 286 electoral votes. Trump's numbers are only totaled through the states he would need in order to get to 270. In those cases, Biden's number is on the left and Trump's is on the right in bold italics.

3 Pennsylvania
 is the state where Biden crosses the 270 electoral vote threshold to win the presidential election, the tipping point state. The tipping point cell is shaded in yellow to denote that and the font color is adjusted to attempt to reflect the category in which the state is.

This group of polls had some impact (and some of it because of an overall lack of polling in some of those Deep South states). But overall, the influence of these polls added today was pretty minimal. Yes, Louisiana shifted on the map and moved up a few cells on the Spectrum just inside the upper end of the Lean Trump category. Alabama also moved up a few spots in the order to the top of the far right column on the Spectrum. As noted above, however, the Yellowhammer state remains a Strong Trump state. No other state moved any on the Spectrum, but ME CD1 shifted below rather than above Maryland in the order. It was a change, but a subtle one.

Where the most alteration occurred between Sunday and Monday was on the Watch List. Three states  -- Arkansas, Louisiana and Texas -- rejoined the group of states within a point of changing categories here at FHQ. While there were polls in Louisiana and Texas to trigger those changes, there was nothing new out of Arkansas. The lone poll in the Natural state earlier in the summer means that the averages in Arkansas are still tethered to other states around which it finished in 2016. And the drop in the average in Alabama based on the Tyson poll influenced the projected margin in Arkansas, moving it back on the Watch List just above the Strong/Lean line.



Where things stood at FHQ on September 21 (or close to it) in...
2016
2012
2008



--
NOTE: Distinctions are made between states based on how much they favor one candidate or another. States with a margin greater than 10 percent between Biden and Trump are "Strong" states. Those with a margin of 5 to 10 percent "Lean" toward one of the two (presumptive) nominees. Finally, states with a spread in the graduated weighted averages of both the candidates' shares of polling support less than 5 percent are "Toss Up" states. The darker a state is shaded in any of the figures here, the more strongly it is aligned with one of the candidates. Not all states along or near the boundaries between categories are close to pushing over into a neighboring group. Those most likely to switch -- those within a percentage point of the various lines of demarcation -- are included on the Watch List below.

The Watch List1
State
Potential Switch
Arkansas
from Strong Trump
to Lean Trump
Georgia
from Toss Up Trump
to Toss Up Biden
Iowa
from Toss Up Trump
to Toss Up Biden
Louisiana
from Lean Trump
to Strong Trump
Mississippi
from Strong Trump
to Lean Trump
Nevada
from Toss Up Biden
to Lean Biden
Ohio
from Toss Up Trump
to Toss Up Biden
Pennsylvania
from Lean Biden
to Toss Up Biden
South Carolina
from Lean Trump
to Toss Up Trump
Texas
from Toss Up Trump
to Toss Up Biden
1 Graduated weighted average margin within a fraction of a point of changing categories.

--
Methodological Note: In past years, FHQ has tried some different ways of dealing with states with no polls or just one poll in the early rounds of these projections. It does help that the least polled states are often the least competitive. The only shortcoming is that those states may be a little off in the order in the Spectrum. In earlier cycles, a simple average of the state's three previous cycles has been used. But in 2016, FHQ strayed from that and constructed an average swing from 2012 to 2016 that was applied to states. That method, however, did little to prevent anomalies like the Kansas poll that had Clinton ahead from biasing the averages. In 2016, the early average swing in the aggregate was  too small to make much difference anyway. For 2020, FHQ has utilized an average swing among states that were around a little polled state in the rank ordering on election day in 2016. If there is just one poll in Delaware in 2020, for example, then maybe it is reasonable to account for what the comparatively greater amount of polling tells us about the changes in Connecticut, New Jersey and New Mexico. Or perhaps the polling in Iowa, Mississippi and South Carolina so far tells us a bit about what may be happening in Alaska where no public polling has been released. That will hopefully work a bit better than the overall average that may end up a bit more muted.


--
Related posts:
The Electoral College Map (9/20/20)

The Electoral College Map (9/19/20)

The Electoral College Map (9/18/20)


Follow FHQ on TwitterInstagram and Facebook or subscribe by Email.

Sunday, September 20, 2020

The Electoral College Map (9/20/20)

Update for September 20.


As the weekend comes a close with just 44 days until November 3, there was another round of polls released. Seven new surveys from six states -- three current Biden states and three Trump states -- that are Lean or closer in the FHQ averages had something good for both candidates.


Polling Quick Hits:
Florida
(Biden 48, Trump 46)
[Current FHQ margin: Biden +3.74]
The Sunshine state continues to be probably the clearest example of a narrowing race. Florida still favors Biden, but that advantage has drawn closer and closer over time, and the new YouGov survey did little to alter that trajectory. If anything, the trend line in the YouGov series in Florida is a microcosm of that change. The firm was last in the field in the Sunshine state in early July and had Biden pulling in the same 48 percent support. Trump, however, has made up ground since that time, adding four percent to bring him within two.


Georgia
(Trump 46, Biden 45)
[Current FHQ margin: Trump +0.38]
Redfield & Wilton Strategies may not have conducted a survey in the Peach state until now, but their findings in this first poll there nearly match the established average shares of support both Biden and Trump hold at FHQ. Trump's 46 in the poll is the same as his average share of support while Biden would round up to 46 from 45.9. So it is not that far off for starters, but also has little impact on the average margin. It nudges it up in Trump's direction by the slightest amount, but keeps Georgia as the most competitive state on the president's side of the partisan line. Georgia may or may not turn blue in November, but that it is the closest state on the board says a lot about where this race stands. If Georgia remains in that position on election day, then Biden is likely going to be in good shape if the order of states holds up and the same basic uniform swing persists.


Minnesota
(Biden 51, Trump 42)
[Current FHQ margin: Biden +8.25]
Redfield & Wilton Strategies also had not been in the field in the North Star state. But there, too, the firm shows a race that is remarkably consistent with the existing averages (candidate shares of support and margin) at FHQ. The 51-42 Biden lead in this survey mirrors the average 51-42 lead the former vice president has here. It is also in line with the bulk of September polling in the the Land of 10,000 Lakes. Again, as with Georgia above, such a survey does little to change the outlook in a state that has been a flip target of the Trump campaign.


Montana
(Trump 49, Biden 42)
[Current FHQ margin: Trump +7.71]
This was the first Siena/NYT Upshot_ survey of the Treasure state, so there is no direct comparison, but it is another poll that tracks fairly closely to the existing FHQ averages. After rounding, Trump leads here 50-43. And while that is a good sign for the president, it comes with a significant caveat. Montana was a state the president carried by 20 points four years ago. And while he is unlikely to relinquish the state to Biden, the 13 point swing in the Democrats' direction is notable. In calendar 2020 polling in the state, the president has run nearly six points behind his 2016 pace, just barely cracking 50 percent. Biden, meanwhile, has boosted the Democrats' share of support by almost eight points.


Pennsylvania
(Biden 48, Trump 43 via Climate Nexus | Biden 47, Trump 45 via Trafalgar Group)
[Current FHQ margin: Biden +5.20]
The Keystone state remains the tipping point state in the order of states depicted in the Electoral College Spectrum below and is just on the Lean side of the five point line separating the Biden Lean and Toss Up states. The Climate Nexus survey may be more in line with that, but the Trafalgar poll has the advantage of being the second poll in a series. And the early July release had Biden up by a margin consistent with the current average margin during Biden's peak period in 2020 polling. But that five point edge for the former vice president in June/July has shrunk to just two points now at Trafalgar. Unlike, say, Florida above, however, Pennsylvania has not seen the same type of narrowing of late. Mid-single digit margins in Biden's favor have been more prevalent in Pennsylvania polling in September.


Texas
(Trump 48, Biden 46)
[Current FHQ margin: Trump +1.14]
Finally, another YouGov battleground tracker continues to show a close race for Texas' 38 electoral votes. The Lone Star state continues to tilt in Trump's direction, but it also continues to be much closer than it has in more than a generation. But the persistence of that narrow advantage for the president there can be seen in the two polls conducted by YouGov in Texas. The one point lead Trump had in the July poll has bumped up to two in September with the president trending toward 50 percent. No, that is not indicative of any tightening, but it is evidence that some wavering support has come home to the president in Texas.






NOTE: A description of the methodology behind the graduated weighted average of 2020 state-level polling that FHQ uses for these projections can be found here.


The Electoral College Spectrum1
DC-3
MA-11
(14)2
CT-7
(162)
WI-10
(253)
AK-3
(125)
IN-11
(60)
HI-4
(18)
NJ-14
(176)
PA-203
(273 | 285)
SC-9
(122)
UT-6
(49)
CA-55
(73)
OR-7
(183)
NV-6
(279 | 265)
MO-10
(113)
KY-8
(43)
VT-3
(76)
NM-5
(188)
FL-29
(308 | 259)
MT-3
(103)
AL-9
(35)
NY-29
(105)
VA-13
(201)
AZ-11
ME CD2-1
(320 | 230)
KS-6
NE CD1-1
(100)
ID-4
(26)
WA-12
(117)
CO-9
(210)
NC-15
(335 | 218)
MS-6
(93)
ND-3
(22)
ME CD1-1
MD-10
(128)
ME-2
(212)
GA-16
(203)
AR-6
(87)
SD-3
(19)
IL-20
(148)
MN-10
(222)
OH-18
(187)
NE-2
(81)
OK-7
(16)
RI-4
(152)
MI-16
(238)
IA-6
(169)
LA-8
(79)
WV-5
(9)
DE-3
(155)
NE CD2-1
NH-4
(243)
TX-38
(163)
TN-11
(71)
WY-3
NE CD3-1
(4)
1 Follow the link for a detailed explanation on how to read the Electoral College Spectrum.

2 The numbers in the parentheses refer to the number of electoral votes a candidate would have if he or she won all the states ranked prior to that state. If, for example, Trump won all the states up to and including Pennsylvania (Biden's toss up states plus the Pennsylvania), he would have 286 electoral votes. Trump's numbers are only totaled through the states he would need in order to get to 270. In those cases, Biden's number is on the left and Trump's is on the right in bold italics.

3 Pennsylvania
 is the state where Biden crosses the 270 electoral vote threshold to win the presidential election, the tipping point state. The tipping point cell is shaded in yellow to denote that and the font color is adjusted to attempt to reflect the category in which the state is.

Three red state polls and three blue state polls did little to alter the outlook here at FHQ despite coming from a host of states that are among the most highly targeted. In most cases, the polls ended up closely mirroring the existing graduated weighted averages. The map remained the same and the Spectrum and Watch List did as well. The race will enter the new work week -- just six weeks out from election day -- looking much as it did when the last week began.



Where things stood at FHQ on September 20 (or close to it) in...
2016
2012
2008



--
NOTE: Distinctions are made between states based on how much they favor one candidate or another. States with a margin greater than 10 percent between Biden and Trump are "Strong" states. Those with a margin of 5 to 10 percent "Lean" toward one of the two (presumptive) nominees. Finally, states with a spread in the graduated weighted averages of both the candidates' shares of polling support less than 5 percent are "Toss Up" states. The darker a state is shaded in any of the figures here, the more strongly it is aligned with one of the candidates. Not all states along or near the boundaries between categories are close to pushing over into a neighboring group. Those most likely to switch -- those within a percentage point of the various lines of demarcation -- are included on the Watch List below.

The Watch List1
State
Potential Switch
Georgia
from Toss Up Trump
to Toss Up Biden
Iowa
from Toss Up Trump
to Toss Up Biden
Mississippi
from Strong Trump
to Lean Trump
Nevada
from Toss Up Biden
to Lean Biden
Ohio
from Toss Up Trump
to Toss Up Biden
Pennsylvania
from Lean Biden
to Toss Up Biden
South Carolina
from Lean Trump
to Toss Up Trump
1 Graduated weighted average margin within a fraction of a point of changing categories.

--
Methodological Note: In past years, FHQ has tried some different ways of dealing with states with no polls or just one poll in the early rounds of these projections. It does help that the least polled states are often the least competitive. The only shortcoming is that those states may be a little off in the order in the Spectrum. In earlier cycles, a simple average of the state's three previous cycles has been used. But in 2016, FHQ strayed from that and constructed an average swing from 2012 to 2016 that was applied to states. That method, however, did little to prevent anomalies like the Kansas poll that had Clinton ahead from biasing the averages. In 2016, the early average swing in the aggregate was  too small to make much difference anyway. For 2020, FHQ has utilized an average swing among states that were around a little polled state in the rank ordering on election day in 2016. If there is just one poll in Delaware in 2020, for example, then maybe it is reasonable to account for what the comparatively greater amount of polling tells us about the changes in Connecticut, New Jersey and New Mexico. Or perhaps the polling in Iowa, Mississippi and South Carolina so far tells us a bit about what may be happening in Alaska where no public polling has been released. That will hopefully work a bit better than the overall average that may end up a bit more muted.


--
Related posts:
The Electoral College Map (9/19/20)

The Electoral College Map (9/18/20)

The Electoral College Map (9/17/20)


Follow FHQ on TwitterInstagram and Facebook or subscribe by Email.