“It is the worst campaign I’ve ever seen and I’ve been watching them since 1980. They’re on the wrong issues. They’re on the wrong message. They’ve got their heads up their assess. … Your damn job is to get your candidate to talk about things that are relevant to the people you need to reach. And if you can’t do your damn job then get out.”
- Republican pollster Frank Luntz on the Trump campaign, quoted in The Hill
One thing that people misremember about 2016 is that there were warning signs that Clinton was in trouble in late October of that year. To be fair, most people do not actually “misremember” this because they never heard about the warning signs in the first place. Polling data in Congressional districts in mid to late October 2016 showed movement to Trump that the national polling was not picking up - particularly the increase in share of vote among white voters. According to Dan Wasserman of the Cook Political Report, “In late October of 2016, district-level polling was full of flashing red warning signs for Hillary Clinton: from northern Wisconsin to New York's Southern Tier and Michigan's Upper Peninsula, single-digit leads for Donald Trump in September had expanded to double digits in heavily white working-class areas. And although much of this party polling was never released publicly, it turned out to be prescient.”
We are now hitting late October 2020 and there is polling data at the Congressional district level. Much of this data has not been not released publicly, but it’s Wasserman’s job to know what’s going on: “[D]istrict-level polls are full of danger signs for Trump. In both parties' private surveys — used to make key resource allocation decisions — he's routinely underperforming his 2016 margins by eight to ten points, consistent with national polls. As a result, one well-placed GOP member told us this week ‘it would be a pleasant surprise if we only lost ten House seats.’"
Trump is doing as well or better in districts with significant Latino populations in South Texas, South Florida, and California’s Central Valley. But, he’s losing – in some cases by double-digits – in many suburban districts he won in 2016. According to Wasserman: “For example, in 2016 Trump carried the districts held by Reps. Dave Schweikert (AZ-06), Susan Brooks (IN-05), Ann Wagner (MO-02) and Chip Roy (TX-21) by around ten points. Today, there is widespread agreement among party strategists that Trump is trailing in those seats, in some cases outside the margin of error. The resulting drag may be too much for down-ballot Republicans to overcome.”
As a result of the latest information, CPR has changed their ratings for 12 Congressional districts. All but one go in favor of the Democrats. However that seat (IL-17) is still Lean Democrat.
CA-50: VACANT (Hunter) (R) - Likely R to Lean R
FL-26: Debbie Mucarsel-Powelll (D) - Toss Up to Lean D
GA-06: Lucy McBath (D) - Lean D to Likely D
IL-13: Rodney Davis (R) - Lean R to Toss Up
IL-17: Cheri Bustos (D) - Likely D to Lean D
KS-03: Sharice Davids (D) - Likely D to Solid D
MI-03: (OPEN) Amash (I) - Lean R to Toss Up
MN-01: Jim Hagedorn (R) - Lean R to Toss Up
MN-02: Angie Craig (D) - Lean D to Likely D
NC-09: Dan Bishop (R) - Likely R to Lean R
VA-02: Elaine Luria (D) - Toss Up to Lean D
WA-03: Jaime Herrera Beutler (R) - Likely R to Lean R
CPR projects a 5-15 seat gain for the Democratic majority. If the trends seen in the district level polling continue, the gain could be larger.
Polling average drop for Biden: Should we be concerned?
Biden’s advantage in the FiveThirtyEight polling average has dropped to plus 10.0 points today. That is likely due to how close today’s IBD/TIPP (A/B) tracking poll is (Biden +3) and a new Rasmussen poll showing a similar margin. However, YouGov (B), SurveyUSA (A), and the USC Dornsife (B/C) tracking poll all report margins around plus 10.
The news about the district level polling is consistent with Biden maintaining a lead of about ten points. The race could narrow, but there is still plenty of room for error for Team Biden. There’s no room for error for Team Trump. Aside from Rasmussen, Trump’s approval rating is in the low 40s among nearly all pollsters. Approval ratings track closely to voter preference. Gallup shows that incumbent presidents do not reelection with low approval ratings. Only George W. Bush was reelected with a final approval rate of under 50%, but he was over 50% in the weeks prior to the election.
Trump’s approval ratings along with the district level polling and Biden’s 10-point polling average lead provide us with a kind of consilience of data all pointing towards a Democratic victory on November 3rd.
So goes Iowa, so goes the nation?
Siena/NYT (A+) has released a poll this morning that has Biden up by three points in Iowa. Why is Iowa important? It only has 6 electoral votes. Mostly because it is a bellwether for a bad night for the GOP if it goes Dem. Also because Trump needs every single electoral vote he can get. Where is he going to replace six votes? New Hampshire and Maine? Not likely. Nevada? Also not likely.
Chances are that if Iowa goes for Biden, there is no way that Trump can win the election. However, the evening could be over before then if Florida (and its 29 electoral votes) goes for Biden. Trump has no plausible route to victory without Florida. Consider this map (made by the interactive tool at Electoral Vote Map):
Biden can lose Pennsylvania (20 EVs), Florida (29 EVs), and Maine-02 (1 EV) and still win exactly the 270 electoral votes needed to win. If Trump loses Iowa, where can he pick up the lost votes? I don’t see a realistic opportunity for him to do so. In any case, the current polling suggests Biden is winning ME-02 and Pennsylvania. Florida, unsurprisingly, is a Toss-Up.
This is what the election will look like if all the Toss-Up states go to Trump, but Biden wins the states he’s ahead in the polling right now:
As you can see, the tipping state is Pennsylvania. Assuming no significant changes in favor of Trump between now and November 3rd, if Pennsylvania is called for Biden the election is over. Even if ME-02 goes for Trump (and current polling indicates that is unlikely - but not implausible), Biden still wins 291 electoral votes. It does not seem plausible at this point that Biden would win Pennsylvania and lose Michigan or Wisconsin. Pennsylvania appears to be the one state of the three that Trump might have a shot at, and for that reason don’t be surprised by a lot of election chicanery from Republicans in the Keystone State on Election Day.