|
Post by mhbruin on Oct 4, 2020 8:12:50 GMT -8
The number on leads in the polls are from 538. The win percentages are from The Economist.
| EVs | BIDEN LEAD Oct 4 | BIDEN LEAD Sept 23 | BIDEN LEAD
Sept 12 | BIDEN LEAD Sept 3 | BIDEN WIN % Oct 4 | BIDEN WIN% Sept 23 | BIDEN WIN %
Sept 12 | BIDEN WIN % Sept 3 | Arizona | 11 | 3.4% | 3.6% | 5.0% | 4.6% | 59% | 60% | 62%
| 65% | Florida | 29 | 2.9% | 1.9% | 2.7% | 4.1% | 69% | 66% | 72% | 68% | Georgia | 16 | 0.6% | -0.6% | -1.5% | -1.5% | 47% | 40% | 42% | 39% | Iowa | 6 | -1.4% | -0.6% | -1.6% | -1.8% | 33% | 34% | 30% | 30% | Maine | 4 | 15.0% | 16.6% | 10.9% | 9.7% | 99% | 98% | 97% | 94% | Michigan | 16 | 6.9% | 7.4% | 7.6% | 6.4% | 90% | 89% | 88% | 82% | Minnesota | 10 | 9.1% | 9.5% | 7.4% | 6.0% | 93% | 92% | 89% | 88% | Nevada | 6 | 6.0% | 5.5% | 5.9% | 6.4% | 84% | 81% | 85% | 85% | New Hampshire | 4 | 9.5% | 9.1% | 6.8% | 7.9% | 88% | 87% | 87% | 84% | North Carolina | 15 | 1.4% | 1.4% | 1.4% | 1.8% | 58% | 59% | 54% | 51% | Ohio | 18 | 0.7% | -0.6% | -0.9% | -1.9% | 37% | 34% | 32% | 31% | Pennsylvania | 20 | 6.1% | 4.7% | 5.1% | 5.6% | 84% | 84% | 80% | 77% | Texas | 38 | -2.4% | -1.2% | -0.8% | -1.5% | 19% | 27% | 31% | 32% | Virginia | 13 | 11.3% | 10.8% | 10.8% | 9.7% | 96% | 97% | 95% | 94% | Wisconsin | 10 | 6.6% | 6.6% | 6.5% | 5.4% | 88% | 87% | 85% | 81% |
Without these swing states Biden has 199 EVs. If we give him likely wins in New Hampshire, Minnesota, and Virginia, he has 232. In order, these are the most likely Biden wins: He needs 38 more. Michigan 16 (6.9%, 90%) Wisconsin 10 (6.6%, 88%) Pennsylvania 20 (6.1%, 84%) Nevada 6 (6.0%, 84%) Arizona 11 (3.4%, 59%) Florida 29 (2.9%, 69%)
You can see different combinations to get to 38, although the most likely is still Michigan, Wisconsin, and Pennsylvania. The big takeaways? 1) At the state level the race is mostly staying stable just as at the national level. The longer it stays this way, the better for Biden, since he is ahead. And the votes are pouring in. 2) Most of the movements are well within the margin of error, but Georgia definitely seems to be moving Biden's way. 3) There is a lot of crappy polling of Ohio. The quality polling shows Biden well ahead. Ohio could be decided on November 3rd, since they count votes as they come in. A Biden win in Ohio decides the election. 4) Florida, Florida, Florida. It's the one state where 538 and The Economist don't seem in total synch.
|
|
|
Post by mhbruin on Oct 4, 2020 9:34:33 GMT -8
Meanwhile Biden's lead is up to 8% in the 538 average, thanks in part to an A- poll showing Biden up by 14%.
That is the biggest lead reported by any highly-rated pollster since July.
|
|
|
Post by mhbruin on Oct 4, 2020 9:56:34 GMT -8
From that highly-rated poll:
The biggest declines for Trump in the poll come from seniors (who are now backing Biden by a 62 percent-to-35 percent margin) and suburban women (58 percent to 33 percent).
And men 50 years and older moved from a 13-point advantage for Trump in the pre-debate NBC News/WSJ poll, to a 1-point advantage for Biden in this latest poll.
|
|
|
Post by mhbruin on Oct 4, 2020 20:24:40 GMT -8
Then there is this from G. Elliott Morris: Trump's polls take a dive after an ugly debate and positive covid-19 test
Voters lack sympathy for the president, and his diagnosis may even be changing a few minds about November
A number of polls were released this weekend, and all of them contained bad news for the president in one way or another. Let’s start with the most germane questions about his covid-19 diagnosis.
A poll from ABC News and Ipsos conducted throughout Friday and Saturday had the starkest findings for Trump. It found that 72% of voters thought that Trump did not take the "risk of contracting the virus seriously enough" or take "the appropriate precautions when it came to his personal health." That includes 43% of Republicans and 95% of Democrats — a notably substantial rebuke of the president from his own party.
A broader consequence of Trump’s diagnosis is that voters have gotten slightly more worried about contracting covid-19. That’s especially true for Republicans — which speaks more than a little to how partisans will follow their opinion leaders until a really shocking event shakes their motivated reasoning. According to the ABC/Ipsos poll, 81% of Americans are either very or somewhat concerned about getting the virus, up from 72% two weeks ago. The share of Republicans who are concerned rose 18 points from 52% to 70%, and among independents, the share rose from 69% to 82%.
Yet there is only mixed evidence that Trump’s covid diagnosis has hurt him all that much in pre-election polling. A separate poll from Ipsos and Reuters conducted between Friday and Saturday showed Biden with a 10-point margin nationally, up from a nine percentage point margin in their poll taken throughout Wednesday and Thursday. We also have some anecdotal data and off-the-record reports from private pollsters that things have shifted modestly in Biden’s direction over the last few days. But we need more data to be sure.
This bad news dovetails with substantial electoral fallout from the debate in the week. As I detailed in a subscribers-only post, most of the scientific polling conducted immediately after the debate found markedly negative reactions to the president’s brash debate style, which included a hitherto-unforeseen pace of interruptions and degree of disrespect for decorum. Accordingly, a poll from Yahoo and YouGov released Sunday found the president’s support slipping 4 points, which is in the margin of error for their survey but matches the trends in other polls. Yahoo reports that among registered voters who watched the raucous clash, 46% said Biden won; only 28% said the same of Trump. The president’s hectoring performance — a clear majority of debate watchers (59%) rated it as “poor,” compared to only 32% for Biden — was followed by a collapse in his support among likely independent voters. Before the debate, these independents narrowly preferred Trump (44%) over Biden (43%). After the debate, they preferred Biden (44%) over Trump (29%) — meaning the president effectively lost 15 points among independents overnight.
Similarly, on Sunday, an 800-person post-debate poll from NBC News and the Wall Street Journal (but fielded before news broke that Trump had covid) gave Biden his best margin of the campaign so far. They find that Biden is now ahead of Trump by 14 percentage points, 53% to 39% among registered voters — up from an 8-point lead in their pre-debate poll. (NB: That 14% is approaching outlier territory, but the trend is pointing in a similar direction to other polling.)
The NBC/WSJ poll also registered an increase in Joe Biden’s favorability ratings, which might help to explain how he has managed to stay above 50% in good polling averages over the election season:
To me one of the more interesting timelines in latest NBC/WSJ poll is Biden’s fav/unfav: 1/20 = -10 8/20= -6 9/20 = -2 10/20 = +2 The attacks on ‘socialism’ and ‘sleepy’ have not only NOT moved numbers away from Biden, but Biden’s actually gotten more popular since the summer.
Another issue that might have hurt Trump over the last week is his nominating Amy Coney Barrett to the Supreme Court. According to NBC/WaPo, 50% of registered voters say they would prefer if the Senate waited to fill the seat left vacant by Ruth Bader Ginsburg until there's a winner in the presidential election. In comparison, 38% want the Senate to hold hearings and a vote before the election.
Finally, I want to draw your attention to two polls from YouGov and CBS News: one in Ohio and one in Pennsylvania. They find Biden and Trump tied in the Buckeye State and Biden up 7 among Keystone-state voters. Those are remarkably similar to the various polling averages and models in both states and indicative of some pretty significant shifts toward Biden there over the last few weeks. Regarding Pennsylvania, which is the likeliest tipping-point state, no high-quality poll (be it live phone or online) has found Biden with a lead smaller than 4 points over the last week. A poll from the New York Times on Saturday also clocked in at +7 for Biden.
Overall, as my analysis of the presidential race goes, I ended this week roughly where I began. The president cannot catch a break in the news cycle, which has been titled against him for weeks on end. That is, of course, a huge barrier to making any progress against both his 8 percentage point national deficit and a slightly smaller hole in the key states.
With only 30 days left, and this last week being what it was, it looks increasingly unlikely that Trump can do what it takes to win the election. Each of my toy election model, the Economist forecast, the FiveThirtyEight model and New York Times polling averages have slid against him over the last few days. Of course, that’s not to say that he can’t win, but his chances of doing so are getting grimmer by the day.
|
|
dsc
Resident Member
Posts: 759
|
Post by dsc on Oct 4, 2020 20:47:11 GMT -8
From that highly-rated poll: The biggest declines for Trump in the poll come from seniors (who are now backing Biden by a 62 percent-to-35 percent margin) and suburban women (58 percent to 33 percent). And men 50 years and older moved from a 13-point advantage for Trump in the pre-debate NBC News/WSJ poll, to a 1-point advantage for Biden in this latest poll. None of these is making sense. Trump's approval has been inching up in the last two months and now is back up to 44% which matches his previous high from the Covid bump in late March. With massive swings in the two key demographics that make up Trump's base (seniors and men), why are his approval numbers not only holding up, but also bumping up recently? Voters who approve of the incumbent do not like to switch horses. Is there a significant number of voters who approve Trump's job performance AND plan to vote for Biden? projects.fivethirtyeight.com/trump-approval-ratings/
|
|
|
Post by northbruin40 on Oct 5, 2020 0:07:24 GMT -8
None of these is making sense. Trump's approval has been inching up in the last two months and now is back up to 44% which matches his previous high from the Covid bump in late March. With massive swings in the two key demographics that make up Trump's base (seniors and men), why are his approval numbers not only holding up, but also bumping up recently? Voters who approve of the incumbent do not like to switch horses. Is there a significant number of voters who approve Trump's job performance AND plan to vote for Biden? projects.fivethirtyeight.com/trump-approval-ratings/I've noticed that too. The best explanation I can come up with is who are doing the popularity polls are more friendly to Trump. In the 538 popularity poll listings I could find 63 polls that were taken in the latter half of September to present. Only 18 of those polls show Trump with a popularity above 44%. Those are (sorry for the formatting) Date ......... Poll ........Popularity ... unPopularity Oct. 2 John Zogby Strategies/EMI Research Solutions 50% 47% SEP. 29-OCT. 1 Rasmussen Reports/Pulse Opinion Research 46% 53% SEP. 26-30 YouGov 46% 53% SEP. 24-28 YouGov 47% 51% SEP. 24-28 Rasmussen Reports/Pulse Opinion Research 47% 52% SEP. 14-28 Gallup 46% 52% SEP. 23-27 McLaughlin & Associates 49% 51% SEP. 22-24 Harris Poll 47% 53% SEP. 22-24 Siena College/The New York Times Upshot 46% 50% SEP. 20-24 YouGov 45% 51% SEP. 22-23 Emerson College 46% 50% SEP. 21-23 Rasmussen Reports/Pulse Opinion Research 52% 48% SEP. 19-22 YouGov 45% 49% SEP. 18-22 YouGov 47% 48% SEP. 18-21 HarrisX 48% 52% SEP. 16-20 Rasmussen Reports/Pulse Opinion Research 51% 48% SEP. 15-18 YouGov 46% 49% SEP. 13-16 NBC News/The Wall Street Journal 45% 53% A few of those are high quality polls (Siena/NYT, Emerson, NBC), and some are very frequent polls (You Gov). Most YouGov polls have smaller numbers for popularity and aren't shown here. But Zogby, Rasmussen, Gallup, McLaughlin and Harris are known Republican friendly polls. Rasmussen, in particular, has been crazy the the last few months. I suspect it is sometimes deliberately biased to favor Trump to satisfy clients/friends ... and other times is adjusted to normal to give them a measure of "truth". Perhaps it's better to use the median rather than the mean, but I'm too lazy to calculate it. OK, so I did calculate the median for the last 10 polls and it's 43% That compares to 42.8% which is Trump's number in the latest 538 polls for the election. Edit: Several Trump friendly popularity polls came out in the last few days of September that did not correspond to changes in the Biden-Trump polling. Those included one by YouGov ... and the others were McLaughlin & Associates, Gallup, and Rasmussen Reports.
|
|
|
Post by mhbruin on Oct 5, 2020 8:19:09 GMT -8
None of these is making sense. Trump's approval has been inching up in the last two months and now is back up to 44% which matches his previous high from the Covid bump in late March. With massive swings in the two key demographics that make up Trump's base (seniors and men), why are his approval numbers not only holding up, but also bumping up recently? Voters who approve of the incumbent do not like to switch horses. Is there a significant number of voters who approve Trump's job performance AND plan to vote for Biden? projects.fivethirtyeight.com/trump-approval-ratings/This is just a guess. Job approval is probably polled among all adults. Election polls at this point are done for likely voters. The weightings used in those types of polls are going to be different It is quite possible that disengaged people approve of the President, mostly because they don't pay attention to the news, and these people are not likely to vote. Again, just guesses on my part. OTOH, the amount of polling showing Biden leading and the lead seeming to grow seems pretty overwhelming.
|
|
dsc
Resident Member
Posts: 759
|
Post by dsc on Oct 5, 2020 8:41:13 GMT -8
None of these is making sense. Trump's approval has been inching up in the last two months and now is back up to 44% which matches his previous high from the Covid bump in late March. With massive swings in the two key demographics that make up Trump's base (seniors and men), why are his approval numbers not only holding up, but also bumping up recently? Voters who approve of the incumbent do not like to switch horses. Is there a significant number of voters who approve Trump's job performance AND plan to vote for Biden? projects.fivethirtyeight.com/trump-approval-ratings/This is just a guess. Job approval is probably polled among all adults. Election polls at this point are done for likely voters. The weightings used in those types of polls are going to be different It is quite possible that disengaged people approve of the President, mostly because they don't pay attention to the news, and these people are not likely to vote. Again, just guesses on my part. OTOH, the amount of polling showing Biden leading and the lead seeming to grow seems pretty overwhelming. In 2012, a great deal was made of Obama's approval ratings in the weeks leading up to the election. The pollsters and media seemed obsessive about it.
|
|
|
Post by mhbruin on Oct 5, 2020 8:46:42 GMT -8
I think I have read that an incumbent's approval rating is a pretty good predictor of his vote share. If Trump gets 44% of the vote, he is toast.
|
|
|
Post by mhbruin on Oct 5, 2020 9:17:49 GMT -8
PublicPolicyPolling
@ppppolls
·
Oct 4
We've done about 70 private polls this week and a very noticeable trend we're seeing especially the last few days is that where before Biden was maybe getting 6-8% of 2016 Trump voters, that's pretty consistently starting to be more like 10-12%
Rats leaving the sinking ship?
|
|
|
Post by Born2BBruin on Oct 5, 2020 9:21:32 GMT -8
In 2012, a great deal was made of Obama's approval ratings in the weeks leading up to the election. The pollsters and media seemed obsessive about it. Obama's approval ratings were in the mid to high 40's for most of 2012. He hit a low of 44% in Gallup at the end of August, with a net -3%. He also had a -4% net earlier that month, but never anything higher. The week after his 44% approval rating, it jumped to 50%, and was 52% with a +7% net going into election day. Obama won 51.1% of the popular vote.
|
|
dsc
Resident Member
Posts: 759
|
Post by dsc on Oct 5, 2020 10:52:56 GMT -8
In 2012, a great deal was made of Obama's approval ratings in the weeks leading up to the election. The pollsters and media seemed obsessive about it. Obama's approval ratings were in the mid to high 40's for most of 2012. He hit a low of 44% in Gallup at the end of August, with a net -3%. He also had a -4% net earlier that month, but never anything higher. The week after his 44% approval rating, it jumped to 50%, and was 52% with a +7% net going into election day. Obama won 51.1% of the popular vote. My point is, Trump's approval has been trending UP in the last two months, but yet his poll numbers against Biden have held either steady or gone DOWN recently. How could both be true?
|
|
|
Post by northbruin40 on Oct 5, 2020 11:38:35 GMT -8
My point is, Trump's approval has been trending UP in the last two months, but yet his poll numbers against Biden have held either steady or gone DOWN recently. How could both be true? Has he been trending up in popularity? I still suspect there may be a biased poll impact. But I did go and look at a consistent polling source. Morning Consult is only a B/C rated polling on 538, but it seems to be a consistent source during the time. They have a weekly polling of over 30,000 likely voters, and more frequent polling of a much smaller number of registered voters. Just using the likely voter polling, in the latter part of July Trump was polling at 39% to 40% in popularity. That's risen to 42% in the later half of September. The registered voter polling is noisier, but seems to have the same trending. However, 42% isn't higher than he is getting polling vs. Biden. I suspect there has been some "coming home" of Republicans that helped his popularity. But as to whether he is actually "more popular" than he is polling vs. Biden? - He average about 43% of the polled vote in September - I suspect that's more of the Rasputin, err I mean Rasmussen effect (generous polls). Oh I suspect some of the gain in popularity was real, but that it's a a fluke that it now appears higher than his polling vs. Biden. Much of that was the low-hanging fruit of Republican-friendly voters wanting a reason to support Trump. The last 1% is just bad polling.
|
|
|
Post by mhbruin on Oct 5, 2020 11:50:11 GMT -8
Nate Silver thinks Biden has gained a bit since the debate: Link
|
|
|
Post by Born2BBruin on Oct 5, 2020 12:57:43 GMT -8
My point is, Trump's approval has been trending UP in the last two months, but yet his poll numbers against Biden have held either steady or gone DOWN recently. How could both be true? The lowest recent score on 538's aggregate approval rating was 40.1% on July 28th (it was that low on July 10th also). That appears to be his lowest approval rating of the year. Today's he's at 44.1%. So, we're talking about a difference of 4%, which is barely outside the margin of error, and since the MOE can overlap, no real difference at all. As far as the election polls are concerned, 538 has trump at 42.7% today. That's a difference of 1.4% from his current approval rating, so no statistical difference. Election polling for trump on July 28th was 41.8%, and on July 10th it was 41.1%; so again, not significantly different from his approval rating, and certainly not showing approval is going up while voting preference is going down.
|
|