Thursday, October 15, 2020

The Electoral College Map (10/15/20)

Update for October 15.

Thursday was another day with a ton of new polling data. There were 22 new surveys from 13 states (and the two congressional districts in Maine) in total. But for all the new numbers, there just were not that many changes to go along with them. It is not that this race is not changing. It is. But it is changing with some measure of subtlety. At this point, the battleground and target states are saturated with polls and despite the fact that older polls are discounted in the FHQ formula, that over-saturation of surveys means that it is difficult to move the needle in any marked way. 

It takes a steady stream of surveys with results noticeably different the average margin (or shares of support) to affect things. Take North Carolina and Pennsylvania as examples. The margin in the Tar Heel state last month was tracking down toward Biden +1.25 but has since reversed course and has today surpassed Biden +1.75. And that is due in part to the recent rush of polls out of North Carolina, many of which have the former vice president ahead by four to five points. Those changes have happened fairly rapidly, but FHQ by design is slow to react (and will likely continue to be slow in changing should any new data continue to reflect the recent reality in North Carolina polling). 

Pennsylvania has followed a similar trajectory, but the changes there have taken place more gradually. Once threatening to jump the Lean/Toss Up line into Toss Up Biden territory, a similar but more spaced out group of polls have nudged the margin in the Keystone state back up to around Biden +5.5 with some signs of plateauing there. 

Other sites may have both of these states a bit further into Biden's column than here at FHQ, but those models are designed to be a bit more responsive to changing polling data. The formula at FHQ is put together a bit differently and the numbers reflect that. There is some general skepticism built in here that may admittedly miss a late break ahead of election day, but operates with the assumption that things ultimately regress to the mean. 

In any event, despite all of the methodological differences, the order of states here is still fairly consistent with what it is at other sites and that is especially true among the battleground and target states. 

But enough of all that. On to the polls... 

Polling Quick Hits:
(Biden 49, Trump 47 via Ipsos | Biden 49, Trump 47 via Monmouth | Biden 49, Trump 45 via OH Predictive)
[Current FHQ margin: Biden +3.10]
Two of the three new polls showed little to no difference over the previous surveys in the series. The latest from Ipsos in the Grand Canyon state continued to give Biden a two point cushion (although both candidate gained a point since the last poll) and the former vice president had the exact 49-47 lead in the Monmouth poll which broke a 47-47 tie (in the low turnout model) in the September poll. But where there was some significant ostensible narrowing was in the OH Predictive survey. There the Democratic nominee's ten point lead was more than halved. And although that shift will grab the attention, that last poll serves as an outlier among the other surveys in Arizona at the time. This is more regressing to the mean more than it is actual tightening in this race. Biden is consistent across all three poll -- a little above his FHQ average share of support -- and the president is more consistent in the OH Predictive poll than in the other two. But none are far off and all are consistent with where the battle for Arizona's 11 electoral votes has been: slightly tilted in Biden's direction. 

(Biden 54, Trump 42 via Civiqs | Biden 54, Trump 39 via Keating Research)
[Current FHQ margin: Biden +12.96]
Consistency is also the name of the game in Colorado. Sure, the Centennial state has been surveyed far fewer times than Arizona, but both of these polls point toward a similar conclusion. Keating was last in the field in the state in May and the picture is hardly different now. Biden is still in the mid-50s and Trump in the upper 30s. And the Civiqs survey -- its first in Colorado in calendar 2020 -- does not stray too far from that bottom line. Again, every time a new Colorado survey is released, it is worth pointing out just how foreign such a wide margin is even relative to 2016 (much less any of the other cycles this century). It is a safe state for Democrats this cycle and has been throughout.

(Biden 50, Trump 47 via Ipsos | Biden 47, Trump 40 via Clearview Research)
[Current FHQ margin: Biden +3.37]
Today was another day with a couple of new polls out of the Sunshine state; the third day in a row. FHQ will focus on the Ipsos poll since this was the first Clearview survey of calendar 2020 and had a wider than average margin while it left a large undecided number sitting out there (9 percent) unprompted. In the Ipsos series, however, there were some subtle changes like the Arizona poll from the firm above. Biden tacked on an additional point while Trump gained two of his own. That marginally narrowed the race through the Ipsos lens but brought the latest survey more in line with the graduated weighted average margin in the Sunshine state at FHQ. And the margin has continued to sort of plateau in the three to four point range. There has been some oscillation, but every move toward contraction is met with data that pushes the candidates further apart.

(Biden 46, Trump 46)
[Current FHQ margin: Biden +0.03] 
Even with yesterday's outlier in the rearview and a new survey from Data for Progress showing the major party candidates knotted at 46, things stayed about where they were a day ago. The Peach state is basically tied -- and has been -- but currently remains tipped in the former vice president's direction by the slimmest of margins. Obviously, a tied poll will do little to change that. And this one look exactly like the last DfP poll in Georgia in mid-September: tied at 46.

(Trump 48, Biden 47)
[Current FHQ margin: Trump +1.27]
Data for Progress was also back in the field for the first time since September in Iowa. In this instance, however, there was some change on the margins. Trump inched up a point and Biden added another two to bring him to within one of the president. That change also pulled the DfP series in line with the FHQ average margin in the Hawkeye state. The candidates' shares are also now roughly in line with their FHQ averages as well. Iowa is close in the polls on average, but for every one Biden lead there are probably two or three for Trump and that is what continues to keep the president narrowly ahead in the state. 

(Biden 50, Trump 40)
[Current FHQ margin: Biden +13.60]

Maine CD1
(Biden 54, Trump 37)
[Current FHQ margin: Biden +23.41]

Maine CD2
(Biden 47, Trump 43)
[Current FHQ margin: Biden +1.51]
While the statewide numbers and those in the first congressional district understate Biden's advantage in each, FHQ will once again focus on the data from the second district in this new Pan Atlantic Research survey of the Pine Tree state. This is the first public poll that the firm has conducted in Maine, so there is no natural comparison. But after the latest Critical Insights survey found Trump up eight in the second, Biden's +4 in this survey serves as a bit of a counter. Yet, the race for that single electoral vote in the more rural northern district in Maine remains the jurisdiction closest to the partisan line on the Biden side, but it pushed the margin a little closer to North Carolina's. On the whole, this Pan Atlantic survey is on par with both candidates' shares of support. It find both candidates right in the hearts of their ranges in the second in any event. 

(Biden 48, Trump 42)
[Current FHQ margin: Biden +7.19]
Thursday was another day with another Michigan poll with Biden ahead in the six to nine point range. The latest (although there is one from Civiqs discussed as part of a wave below) was from RMG Research. And those polls are not doing much to change the outlook in the Great Lakes state. That is particularly true of this survey that falls roughly in line with both candidates' average shares of support at FHQ. One place that it does break from some recent polling in the state is that it finds Biden below 50 percent and at the bottom of his recent range of results there. Still, the status quo was maintained in this one. 

North Carolina
(Biden 51, Trump 46)
[Current FHQ margin: Biden +1.79]
FHQ mentioned at the outset today that those polls with Biden four to five points are fueling a push in the margin back in his favor. The Civiqs update in the Tar Heel state added another datapoint to that. But this is an update to a (now) series of polls that found Biden ahead by three back in May among a sample of registered voters in North Carolina. The transition to likely voters now (and time since May) has shifted things in the former vice president's direction. Trump held steady at 46 percent, but Biden jumped up above 50 percent. That 46 percent is in line with Trump's current average level of support in the state, but Biden's 51 percent is out in front of his while being consistent with a marginal rising tide of support for him. There are a few more 47-50s popping up for Biden than the 45-47 range that was the core of the Democratic nominees polling there.

(Biden 46, Trump 43)
[Current FHQ margin: Biden +5.49]
Yesterday Trafalgar charted out a battle in the Keystone state that was within two points and today it is  Insider Advantage finding Biden up just three. Both have Trump hovering around his average share of support but Biden well below his. Both also have a fairly significant share of respondents that fall into the undecided or other category. It has been those types of polls -- those with an undecided share plus other collectively approaching 10 points -- that have tended to be closer not just in the commonwealth but in other states, both battleground and otherwise, as well. 

South Carolina
(Trump 49, Biden 41 via Siena/NYT Upshot | Trump 52, Biden 43 via Data for Progress)
[Current FHQ margin: Trump +6.75]
While the summer saw a host of surveys in the Palmetto state find the race for the state's nine electoral votes in the mid-single digits, the polling has taken a turn in October. Recently polling in South Carolina though sporadic still has begun to show the president out to leads approaching ten points. That includes both surveys released today. The Siena poll is like a lot from the college pollster. It leaves undecideds unprompted which typically means the candidates fall short of their established average shares. That is true in this case. And while the Siena poll lacks a true point of comparison, the Data for Progress survey does not. And that series has shown some real movement since the September poll. Biden held steady at 43 percent, but Trump consolidated support with his share rising by five points as the undecided respondents fell by an equivalent five points. Now, that 52-43 lead for the president helps to stretch the average margin out there but it still has Trump lagging a few points behind his 2016 showing while remaining above 50 percent. Biden may have improved over Clinton's pace from four years ago, then, but that is all for nought given where the president is. Again, South Carolina on the Trump side of the partisan line looks a lot like those Lean Biden states on the other side. 

(Biden 55, Trump 42 via Civiqs | Biden 53, Trump 38 via Roanoke College)
[Current FHQ margin: Biden +12.30]
Like Colorado above, the polling in Virginia does not look like it has in some past cycles during the 21st century. The Old Dominion is not nearly as competitive as it has been and the polling has continually painted that picture in 2020. Biden is comfortably above 50 percent in the averages in the commonwealth at FHQ and neither of today's two polls diverge from that. The Roanoke polls have consistently fallen in that category in three polls since May. And even though the college pollster has had Biden in the low 50s and now right on his average share of support, they have also repeatedly found Trump in the upper 30s below his average share of support. Regardless, like Colorado, Virginia is seemingly comfortably in Biden's column despite some recent assertions from the Trump campaign.

(Biden 47, Trump 45)
[Current FHQ margin: Biden +6.30]
Another Trafalgar Group poll -- this one from Wisconsin -- does not suffer as acutely from the same drawback discussed above in the Pennsylvania discussion. In this instance, the combined undecided/other share is not as large, but the margin is much closer than some other recent public opinion work in the Badger state. It does, however, find Biden on the low end of his range of recent results as Trump is toward the higher end of his. And one could focus on those issues or point toward the fact that in the Trafalgar series in Wisconsin, little has changed since the firm last conducted a poll there in late September. Trump gained a point and that is it. This is yet another story of consistency.

Civiqs (October Rust Belt Rising wave)

Ohio: Trump +3 (Biden +2, Trump +2 since September round)) [Current FHQ margin: Trump +0.59]
Pennsylvania: Biden +7 (Biden +/-0, Trump +/-0)
Wisconsin: Biden +8 (Biden +2, Trump +1)
Michigan: Biden +9 (Biden -1, Trump +1)

Not to give the last Rust Belt series of polls from Civiqs short shrift, but there was not much movement for either candidate among these four Great Lakes states since September. More importantly, perhaps, the order of the states matches the rank order depicted on the Spectrum below. And while the margin in Ohio may be a bit more in Trump's favor than the average margin here at FHQ, the margins in the other three states are maybe tilted a bit more in the other direction. But they do fall in line with where much of recent polling has been in those three blue wall states that Trump flipped in 2016. 

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
(273 | 285)
(279 | 265)
SC -9
(308 | 259)
(319 | 230)
NE CD1-1
ME CD2-1
(335 | 219)
(351 | 203)
ME CD1-1
NE CD2-1
NE CD3-1
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.

Whereas all that polling a day ago yielded one significant change -- Georgia jumping the partisan line into Biden territory -- today's group held the line. Only South Carolina budged, shifting two cells deeper into the Lean Trump group of states and further away from the Lean/Toss Up line to which it had recently been drawn. No longer does South Carolina seem to be inching toward the Watch List (which remains unchanged from yesterday). But the Palmetto state is in a tightly knit group with Alaska and Missouri. As all of those states have dipped into or flirted with the Toss Up category in 2020, FHQ has said that those three were the states where Biden could tack on some additional electoral votes if the bottom truly dropped out on President Trump. That bottom may or may not drop out between now and election day, but Biden's prospects of adding any electoral votes from this trio of states seems dim. The gap between the last Toss Up Trump state (Texas) and the first Lean Trump state (now Missouri) is nearly five points. That is less a gap and more a chasm. If the bottom drops out on Trump, Biden's advances are likely to end at Texas. But given how consistently Iowa, Ohio and Texas have been tilted in Trump's direction those may even be tough tasks for the Biden campaign. Within range, but difficult flips. 

19 days to go.

Where things stood at FHQ on October 15 (or close to it) in...

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
Potential Switch
from Toss Up Biden
to Toss Up Trump
New Hampshire
from Strong Biden
to Lean Biden
New Mexico
from Strong Biden
to Lean Biden
from Toss Up Trump
to Toss Up Biden
from Lean Biden
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:

Follow FHQ on TwitterInstagram and Facebook or subscribe by Email.

No comments: