Read on Twitter
@JoeBiden v. Trump 9/1 update THREAD
Nat'l: Biden +7.3 / Clinton +3.7
Electoral votes: Biden 301 / Clinton 269
Biden at 50% or above in NINE battlegrounds, ZERO for Clinton (highest 45.8%, 13 under 42%)
N.B. This is 30 days after both 2016 conventions vs. 1 week after 2020 RNC
Nat'l: Biden +7.3 / Clinton +3.7
Electoral votes: Biden 301 / Clinton 269
Biden at 50% or above in NINE battlegrounds, ZERO for Clinton (highest 45.8%, 13 under 42%)
N.B. This is 30 days after both 2016 conventions vs. 1 week after 2020 RNC
NOTES (con’t):
1. "Battleground": ≤10% in 2016 results or 2020 polling averages
2. Polls with 3rd party candidates used, when available
3. High/low turnout-hypothesizing versions of polls not used
1. "Battleground": ≤10% in 2016 results or 2020 polling averages
2. Polls with 3rd party candidates used, when available
3. High/low turnout-hypothesizing versions of polls not used
NOTES (con’t):
I'm excluding one A-B rated pollster, Spry (R), for their use of "push" questions (i.e., crafted to be critical of one party), and results that are erratically misaligned with polling averages on that date (pic 2). I am including all other partisan pollsters.
I'm excluding one A-B rated pollster, Spry (R), for their use of "push" questions (i.e., crafted to be critical of one party), and results that are erratically misaligned with polling averages on that date (pic 2). I am including all other partisan pollsters.
Note to those cherry picking individual polls, looking for convention bumps: they might be real, might not, but we've had ZERO live-interview polls, national or in battlegrounds since both ended. We’ll have a much better idea of the post-convention landscape in a couple of weeks.
Polling versus forecasts:
A1. 2016 polling *averages* were ACCURATE (≥5%), nat'l + swing states, except WI/IA
https://www.aapor.org/Education-Resources/Reports/An-Evaluation-of-2016-Election-Polls-in-the-U-S.aspx
Most election *forecasts*—which this thread is NOT—were not. 538's forecast largely matched odds from Trump’s internal pollsters at ~30%.
A1. 2016 polling *averages* were ACCURATE (≥5%), nat'l + swing states, except WI/IA
https://www.aapor.org/Education-Resources/Reports/An-Evaluation-of-2016-Election-Polls-in-the-U-S.aspx
Most election *forecasts*—which this thread is NOT—were not. 538's forecast largely matched odds from Trump’s internal pollsters at ~30%.
A2. Biden battleground margins 2.8% better than Clinton's (pic 1)
BUT average undecideds + 3rd party 7.5% now, about a THIRD of Clinton's 21.5% on 9/1/16
AND Biden’s lead is larger than / equal to ALL remaining undecideds in NINE battlegrounds, ZERO for Clinton (pic 2)
BUT average undecideds + 3rd party 7.5% now, about a THIRD of Clinton's 21.5% on 9/1/16
AND Biden’s lead is larger than / equal to ALL remaining undecideds in NINE battlegrounds, ZERO for Clinton (pic 2)
A2a. That's because 2020 general electorate has more engaged than Election Day 2016, 2012, or even Obama’s first election in 2008 since JANUARY! This question isn't asked very often by pollsters, but 2020 polls that did find 82-89% "very/extremely" interested or motivated.
A3. Biden should be weaker than Clinton, since he's running against an incumbent, who've won re-election 10 of 13 times since 1916 (LBJ, Truman didn't run)
Per @ForecasterEnten, Biden's in the strongest position of any challenger in modern polling https://www.cnn.com/2020/05/31/politics/biden-maintains-strong-position/index.html?utm_medium=social&utm_source=twCNNp&utm_term=image&utm_content=2020-05-31T19%3A03%3A05
Per @ForecasterEnten, Biden's in the strongest position of any challenger in modern polling https://www.cnn.com/2020/05/31/politics/biden-maintains-strong-position/index.html?utm_medium=social&utm_source=twCNNp&utm_term=image&utm_content=2020-05-31T19%3A03%3A05
A3a. Enten further notes that the only incumbent in modern polling to come back after trailing as badly as Trump is now was Truman in 1948. Carter was in a similar position in 1980, and lost by 440 electoral votes.
https://www.cnn.com/2020/06/27/politics/incumbents-elections-polling-analysis/index.html
https://www.cnn.com/2020/06/27/politics/incumbents-elections-polling-analysis/index.html
(A3a) Also, as much as trailing candidates love the Truman story and this famous pic from 1948, polling methods have been fundamentally changed since then, not least by moving away from "quota sampling," and literal door knocking to find respondents:
https://www.math.upenn.edu/~deturck/m170/wk4/lecture/case2.html
https://www.math.upenn.edu/~deturck/m170/wk4/lecture/case2.html
A3b. Biden hasn't trailed in a SINGLE nat'l poll by an A-B rated pollster in 2020. Clinton trailed in TWELVE by 9/1/16. Just four ties for Biden (two for Clinton), and only one w/undecided factor ≥10% (Atlas 2/3: 45/45); in other words, when leaners break, they break for Biden.
A3b. Speaking of breakers: most pollsters automatically include "leaners" in their totals, but three separated that data, and Biden won them 66-33, 75-25, and 50-50. That means Trump’s hope of a late break in his direction is very unlikely without a major shift in voter sentiment
A3b. Additionally, the not-yet-fully released NBC/Wall St Journal poll shows that those who voted 3rd party in 2016 are backing Biden by more than a 2-to-1 margin, 47%-20%. https://twitter.com/kylegriffin1/status/1300434418415398912
A3b. In fact, Biden’s lead is STEADIEST in modern polling. Why? Because, as expressed in a phrase from the campaign, “WE KNOW JOE.” This is what's kept him ahead, despite the onslaught of smears and criticisms coming from the far right and left. https://www.cnn.com/2020/05/10/politics/biden-polling/index.html
A4. Past "Dem/GOP states" more competitive now, becoming toss-ups (WI, AZ) or leaners (ME, MO).
Partly demographic changes, partly sentiment. Point is, 2016/20 comparisons never exactly equal
https://www.newsweek.com/us-state-population-electoral-college-2020-census-elections-1481019 https://fivethirtyeight.com/features/americas-electoral-map-is-changing/
Partly demographic changes, partly sentiment. Point is, 2016/20 comparisons never exactly equal
https://www.newsweek.com/us-state-population-electoral-college-2020-census-elections-1481019 https://fivethirtyeight.com/features/americas-electoral-map-is-changing/
Here’s a chart that give a general idea of how the battlegrounds polling averages compare to nat’l average in 2020:
https://fivethirtyeight.com/features/why-bidens-polling-lead-is-different-from-clintons-in-2016/
https://fivethirtyeight.com/features/why-bidens-polling-lead-is-different-from-clintons-in-2016/
So why all these comparisons to the last election? First, there are some valuable contrasts to be made, especially since Biden’s running against the same opponent.
But mostly, it’s cuz if ya don’t, people who don’t like the data will just lazily say "BUT 2016!!!"
But mostly, it’s cuz if ya don’t, people who don’t like the data will just lazily say "BUT 2016!!!"
B1. How's he doing it? Per A/B-rated nat'l polls (8/1-31) vs 2016 exits, he's outperforming Clinton:
White: +15
White/college: +19
White/no coll: +16
18-29 yrs: +13
Moderate: +20
Over 65: +11
Suburban +15
Men: +8
White: +15
White/college: +19
White/no coll: +16
18-29 yrs: +13
Moderate: +20
Over 65: +11
Suburban +15
Men: +8
B1a. The most noteworthy of these is that Biden is within 5% of Trump with white voters. No Dem's won this group since LBJ in '64. Only others to get within 5% were Carter 76, Clinton 92/96, and Obama 08, the last three Dems to win the presidency. All others lost by 10-20%+.
B1a. Reflective of this, he's also doing much better with white college graduates, a group that’s been a key swing demographic since 1988. More importantly, he’s doing 16% (!!!) better with white voters who didn't go to college, a group that turned out BIG for Trump in 2016.
B1b. Next is being essentially TIED with men. Again, the last three to do it were Carter 76, Clinton 92/96, and Obama 08, all others lost by 10-20%+.
He’s winning another traditional GOP bloc, voters over 65, which all winning Dems have held since Carter.
He’s winning another traditional GOP bloc, voters over 65, which all winning Dems have held since Carter.
B1b. Biden’s also doing historically well in the suburbs, outpacing Clinton by 15%. This is notable, considering Trump’s recent, less-than-subtle appeals to suburbanites, who were a key swing demographic in 2016.
https://www.cnn.com/2020/07/19/politics/joe-biden-donald-trump-suburban-voters-polling/index.html
https://www.cnn.com/2020/07/19/politics/joe-biden-donald-trump-suburban-voters-polling/index.html
B1c. A popular myth these numbers explode is that Biden's struggling with young voters. While pollsters divide age groups differently (e.g., 18-29 vs. 18-34, 30-44 vs. 35-50), he’s currently running 13% BETTER than Clinton with the youngest group, 18-29.
https://fivethirtyeight.com/features/biden-doesnt-really-have-a-young-voters-problem/
https://fivethirtyeight.com/features/biden-doesnt-really-have-a-young-voters-problem/
B1d. Biden is absolutely OWNING the political middle, doing 11% better with independents than Clinton, and 20% (!!!) better with moderates, while matching conservatives. This also helps Biden, based on a recent Gallup poll showing a substantial break toward the left.
B1e. One last myth to put to bed is that Biden struggles with liberals. In fact, he’s outpacing Clinton by 9%.
B1f. One demographic where Biden has a slight weakness is Hispanic and Black voters: Trump’s 8% stronger than 2016 with Hispanics, and a large number of undecideds still remain. Definitely some work to do.
B2. In May, The Hill noted voters who dislike both break HARD for Biden, by margins ranging from 30-50%+.
Trump won them by 10% in '16, but they didn’t break for him until OCTOBER! https://thehill.com/homenews/campaign/497359-polls-show-biden-has-edge-with-voters-who-dont-like-their-choices
Trump won them by 10% in '16, but they didn’t break for him until OCTOBER! https://thehill.com/homenews/campaign/497359-polls-show-biden-has-edge-with-voters-who-dont-like-their-choices
B2. Also in May, CNN polled "double likers" (a group about the same size as "haters"). They favor Trump by ~40, but advantage is Biden's, since likers less likely to be swayed by smears than their counterparts (esp w/Trump’s higher "very unfavorable"). https://www.cnn.com/2020/05/13/politics/voters-biden-trump-2020-analysis/index.html
B3. One reason Biden's doing so well? Voters just LIKE Biden more than Trump:
Biden's currently +2 in un/favorables, and more importantly, that’s 14% BETTER than Trump, whereas Clinton’s rating was only slightly better than Trump's in 2016.
https://fivethirtyeight.com/features/americans-distaste-for-both-trump-and-clinton-is-record-breaking/
Biden's currently +2 in un/favorables, and more importantly, that’s 14% BETTER than Trump, whereas Clinton’s rating was only slightly better than Trump's in 2016.
https://fivethirtyeight.com/features/americans-distaste-for-both-trump-and-clinton-is-record-breaking/
B3a. Also, the percent of respondents who rate Trump "very" or "highly" unfavorable is consistently MUCH worse for Trump, currently 12% higher than Biden!
B3b. This reflects Trump’s job approval, which is the lowest in overall average of any president in modern polling. At this point in his fourth year, Trump's badly trailing Ford’s final numbers, only slightly ahead of GHW Bush, and leading Carter.
B4. Enthusiasm FOR Biden has risen, post-DNC, but Trump still leads. However, enthusiasm is about turning out the party base, and as we saw in the Dem primaries and 2018 midterms, it's the desire to beat Trump that's turned out record numbers.
4a. A recent poll found 80% of Biden voters have a very unfavorable view of Trump, but vice-versa just 53%. Also, voting AGAINST a candidate has become more powerful than voting for in the 21st century. Again, advantage Biden.
https://fivethirtyeight.com/features/why-trump-not-biden-might-have-an-enthusiasm-problem/
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3414607
https://fivethirtyeight.com/features/why-trump-not-biden-might-have-an-enthusiasm-problem/
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3414607
Speaking of turning out, remember that polls can CHANGE, as we see from the polling maps of August 31st, 2016, and Election Day that year. So whatever the polls say, they don’t mean anything unless you show up and VOTE!
So in closing, VOTE VOTE VOTE VOTE VOTE!!!
So in closing, VOTE VOTE VOTE VOTE VOTE!!!
