Polls of black voters are wrong

[Skill and Chance]

I actually agree something doesn't add up here.  Over and over again, we see maybe a slight shift in a PVI sense in plurality-black areas in 2020's elections, but nothing like the 25%R results that show up over and over again in polling. Fetterman roughly matched Biden in Philadelphia, Warnock roughly matched Biden in Atlanta.  It wasn't as impressive as elsewhere in the state, but Beshear did fine in Louisville.   The Southside VA legislative seats only marginally trended R. 

The strongest case for this comes from Louisiana, but 1. Dems didn't really try and 2. it has a large population of observant Catholic black people who are anomalously pro-life (like half of Dems in the Louisiana legislature supported the abortion ban amendment and then it got more votes than Trump in the referendum in 2020).  None of this generalizes anywhere else. 

The places that really have been zooming right in a PVI sense are either plurality-Hispanic, plurality-Asian and/or have a lot of non-Christian religiously observant people.  That I think will continue and may if anything be underestimated.  If 2024 is a narrow Trump PV win as current polling suggests, my guess for surprising result would be New York closer than Texas. 

[ηєω ƒяσηтιєя]

I actually agree something doesn't add up here.  Over and over again, we see maybe a slight shift in a PVI sense in plurality-black areas in 2020's elections, but nothing like the 25%R results that show up over and over again in polling. Fetterman roughly matched Biden in Philadelphia, Warnock roughly matched Biden in Atlanta.  It wasn't as impressive as elsewhere in the state, but Beshear did fine in Louisville.   The Southside VA legislative seats only marginally trended R.  

The strongest case for this comes from Louisiana, but 1. Dems didn't really try and 2. it has a large population of observant Catholic black people who are anomalously pro-life (like half of Dems in the Louisiana legislature supported the abortion ban amendment and then it got more votes than Trump in the referendum in 2020).  None of this generalizes anywhere else.  

The places that really have been zooming right in a PVI sense are either plurality-Hispanic, plurality-Asian and/or have a lot of non-Christian religiously observant people.  That I think will continue and may if anything be underestimated.  If 2024 is a narrow Trump PV win as current polling suggests, my guess for surprising result would be New York closer than Texas.

Which is not happening and would be completely insane.

[Mr.Bakari-Sellers]

I know that's why it's a 303 DPrez, a 222DH and tied S losing WV and maybe picking up TX, that's why we get everyday  tracking polls Trump +4 and Biden +4 because pollster are having a tough time on blk voters but if you go to blue states they are hardly any Maga signs especially in AZ and NV but BLM abd Red states like. TX AND especially FL its Maga signs, I travel all the time between IL and CA, swath in the middle except CO, NV, NM and AZ are like CA and Dallas, Houston area and the rest is more like FL of red states

[The Economy is Getting Worse]

I actually agree something doesn't add up here.  Over and over again, we see maybe a slight shift in a PVI sense in plurality-black areas in 2020's elections, but nothing like the 25%R results that show up over and over again in polling. Fetterman roughly matched Biden in Philadelphia, Warnock roughly matched Biden in Atlanta.  It wasn't as impressive as elsewhere in the state, but Beshear did fine in Louisville.   The Southside VA legislative seats only marginally trended R. 

The strongest case for this comes from Louisiana, but 1. Dems didn't really try and 2. it has a large population of observant Catholic black people who are anomalously pro-life (like half of Dems in the Louisiana legislature supported the abortion ban amendment and then it got more votes than Trump in the referendum in 2020).  None of this generalizes anywhere else. 

The places that really have been zooming right in a PVI sense are either plurality-Hispanic, plurality-Asian and/or have a lot of non-Christian religiously observant people.  That I think will continue and may if anything be underestimated.  If 2024 is a narrow Trump PV win as current polling suggests, my guess for surprising result would be New York closer than Texas. 

On the other hand black turnout is crashing virtually everywhere except for Mississippi 2023 in the Biden era. Even in Georgia 2022 where it dropped less it was probably due to the fact that 3 of the 4 candidates in Gov/Sen were black.

Maybe Trump only gets marginal gains with black voters(16%, up from 12% in 2020 and 8% in 2016), while turnout crashes to 2004 levels?

[Banana Republican]

Maybe Trump only gets marginal gains with black voters(16%, up from 12% in 2020 and 8% in 2016), while turnout crashes to 2004 levels?

Part of the issue here is that you have an incorrect impression of how Dems did with black voters in 2016 and 2020 (and probably also in 2022). Presumably this is because you are probably quoting figures you found in exit polls, but they have similar methodological problems as regular polls (particularly for 2020 when so many votes were by mail and consequently "exit" polls were in reality normal polls).

Better estimates are that Trump got about 6% of the black vote in 2016 and 9% in 2020. Meanwhile, House Republicans got about 12% of the black vote in 2022 (compared to 7% in 2020, 9% in 2018, and 13% in 2016).

https://www.dropbox.com/s/re0gtn1o57fzwp5/Catalist_What_Happened_2022_Public_National_Crosstabs_2023_05_18.xlsx?dl=0

It IS true that the couple points of difference between those different years can make a difference at the margin, particularly if the overall race is close. But regardless, the numbers in 2024 are going to be roughly in this same range also.

[The Economy is Getting Worse]

Maybe Trump only gets marginal gains with black voters(16%, up from 12% in 2020 and 8% in 2016), while turnout crashes to 2004 levels?

Part of the issue here is that you have an incorrect impression of how Dems did with black voters in 2016 and 2020 (and probably also in 2022). Presumably this is because you are probably quoting figures you found in exit polls, but they have similar methodological problems as regular polls (particularly for 2020 when so many votes were by mail and consequently "exit" polls were in reality normal polls).

Better estimates are that Trump got about 6% of the black vote in 2016 and 9% in 2020. Meanwhile, House Republicans got about 12% of the black vote in 2022 (compared to 7% in 2020, 9% in 2018, and 13% in 2016).

https://www.dropbox.com/s/re0gtn1o57fzwp5/Catalist_What_Happened_2022_Public_National_Crosstabs_2023_05_18.xlsx?dl=0

It IS true that the couple points of difference between those different years can make a difference at the margin, particularly if the overall race is close. But regardless, the numbers in 2024 are going to be roughly in this same range also.

Black turnout dropped 20% from 2018 to 2022, which is big. If Trump gains about 8 points in margin with black voters (probably higher due to third parties as well), and black turnout is 20% lower, this helps Trump by about 4 points in the national popular vote and as much as 8 points in Georgia.

[Banana Republican]

Black turnout dropped 20% from 2018 to 2022, which is big.

You are just making things up. The catalist data shows a drop from 12% of the electorate to 10%. First of all, that is vote share, which is an entirely different thing from turnout. Second of all, that is not a 20% decrease (even in vote share, .1/.12 = .83), and both the 12% and the 10% are rounded to the nearest percentage, so in reality the decrease in vote share from 2018 to 2022 may be more like 1% than 2% (depending on the rounding error).


Also, according to Census CPS data:

In 2018:
63.9% of black citizens age 18+ were registered to vote
51.1% voted (with 22.3% not responding)
https://www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-583.html

In 2022:
64.1% of black citizens age 18+ were registered to vote
45.1% voted (with 21.3% not responding)
https://www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-586.html


.451/.511 = .882. So that is not a 20% decrease either. Try 12%. But of course, even with that 12% figure, if you want to say anything sensible about it then you need to compare to the overall decrease in turnout from 2018 to 2022, and you need to also take into account that there is a difference between midterms with Republican and Democratic Presidents in terms of normal partisan turnout patterns.

If Trump gains about 8 points in margin with black voters (probably higher due to third parties as well), and black turnout is 20% lower, this helps Trump by about 4 points in the national popular vote and as much as 8 points in Georgia.

I can say with the same factual basis as you have supporting you that if Leprechauns exist and vote for Biden en masse, this helps Biden and will help him win Georgia.

It is just as relevant as your claim.

[Progressive Pessimist]

I agree. We do need to accept that the days of Obama winning them by 90%+ are gone, but I still see no concrete reason (outside of polling, which isn't being backed up by election results) why Biden is going to do worse than average/Trump better than average. Or even just compared to 2020.

[Mr.Bakari-Sellers]

Trump tax cuts are disaster for blks and Latinos the polls Pred a Red wave in 22 and were wrong, Steve Konaki said the maps are blank on Eday how many times you gonna believe the polls and they were wrong last time blks don't support Trump if you aren't rich

The polls aren't gonna discourage Ds from voting

[Averroës]

The worry for Democrats is the kind of black voter who (A) isn't going to show up for an effectively uncontested primary with minuscule turnout but (B) has been more likely than not to vote for them in the general election.

On the other hand, there very clearly isn't a market for what Biden's primary opponents are selling.