Share this page!

Lakshya Jain Profile picture
Twitter logo
22 Sep, 5 tweets, 1 min read
Some bizarre things:

(1) the reported results are among *adults*, not RVs
(2) Poll doesn't weight by education, vax rate, party, or region.
(3) Adult sample is only Biden +7
(4) "Do you consider yourself a white (Hispanic/Latino) or a black (Hispanic/Latino)?" is a question?
There's very few universes in which a 55% college-educated electorate votes for Glenn Youngkin by 5 points in 2021.

It's also a bit of a stretch for me to think that in a state that was Biden +10, the *likely voter electorate* is Biden +1.
I think what's happened here is fairly simple: this agency does not appear to have the slightest idea as to what they're doing (they're not even listed on 538 as a pollster from what I can find?)
Glenn Youngkin could absolutely win this race. Please do not think I'm saying he can't. But this poll is not data worthy of considering. And given the flaws we've pointed out above and the lack of margin splits by demographic, this is consequently not going in our model.
BTW, point #4 on itself should be a red flag for any poll. Do. Not. Do. This.

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Lakshya Jain

Lakshya Jain Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @lxeagle17

Lakshya Jain Profile picture
24 Sep
If I had to take a crack at the Senate map for next year, this is probably what I think my ratings are *at the moment*. Could argue with some of these, but largely just depends on the national environment.

Also: lean is not safe. Not saying Dems can't win WI or Rs can't win NH. Image
Wisconsin's low vaccination rates give me a lot of pause. Can Barnes stop the driftless from sliding to the right? I don't think so. Can he make enough inroads in Milwaukee's suburbs to counter it? Maybe, but it won't be easy.
North Carolina: if Rs nominate McCrory to go against Beasley, this could end up being a tougher play than the GOP would like and I can see an argument for Ds being favored given that Beasley can drive up Black turnout strongly while maintaining decent margins with educated whites
Read 5 tweets
Lakshya Jain Profile picture
21 Sep
If you can look past the media noise, Biden's disapproval has slowly narrowed of late on 538, from 4.5 points underwater at its worst to 2.3 points as of today (9/21).

Will be interesting to see how this progresses.
Polls are a snapshot in time.

He'll probably bump back into the green at some point IMO. The historical levels of polarization we have today makes his floor significantly higher than people might expect. As campaigning kicks off for 2022, it might stabilize a bit more.
The bigger question here is whether a change like this is evidence of any meaningful, sustained movement back or whether it's just randomness bouncing it around. Good cases for either, but whatever the answer, it'll probably be drowned out by what happens with infra/debt ceiling.
Read 4 tweets
Lakshya Jain Profile picture
20 Sep
The Canadian election is probably a bit more important, but for those curious, @Thorongil16 and I have an update to our #VAGov model with the Washington Post D+3 poll from Friday.

Our forecast is Democrats winning by 7.3, with an 80% chance of victory.
@Thorongil16 This race has stayed remarkably stable, with few real swings in polling. Even the WaPo D+3 poll was really because of an incredibly R-friendly LV screen that appears to have oversampled non-college whites via a new polling technique. The poll of registered voters was D+6.
There is little in my mind that suggests that this race has substantially changed in the last month. If Biden's approval plunges to around ~42% on FiveThirtyEight or something, we could be in for a very real race. But until then, this race remains likely Democratic.
Read 4 tweets
Lakshya Jain Profile picture
20 Sep
It won't matter as much as the generic national political environment will in 2022, but if the GOP reliance on election day votes doesn't let up a little, it's not hard to imagine it being a little costly in some areas where the weather isn't all that great on November 8, 2022.
These types of things are *significantly* less important than the overall national political environment. But let's say a storm hits in an area where a key seat is up for grabs. What then? Not everyone who would have voted will. Rita Hart lost by 6 votes in 2020.
Overall, it's just hard to escape the thought that the GOP have likely cost themselves some close seats somewhere thanks to the lies about mail ballots that they've been pushing among their base. It's just that the ones they do cost will depend on the national environment.
Read 4 tweets
Lakshya Jain Profile picture
17 Sep
This is also one of the more conservative estimates of the college-educated electorate share -- McAuliffe wins college voters by 22 points (similar to what Biden did) but WaPo's LV screen says the share of college voters *drops* by a few points from 2020. Bit skeptical of that.
If anything, I'd have thought the enthusiasm gap would have been shown more in intra-group turnout differential (e.g. college Rs are more likely to turn out than college Dems), but that's not happening in this poll at all!! It's basically a weird electorate screen.
McAuliffe is winning college voters by a greater margin than Biden did. So it's a bit strange to see, then, that the delta between non-college and college voters is *higher* than it was in 2020, and it cuts counter to most of what we already know re: turnout, hence my skepticism.
Read 5 tweets
Lakshya Jain Profile picture
17 Sep
We've updated our 2021 #VAGov model with the recent Emerson poll. The projection now has McAuliffe (D) winning by 7.5 points, with an 80% chance of victory in November.

You can view county margins for the model made by me and @Thorongil16 at striderforecasts.com
I do expect this to actually move as more polling comes in, but we're currently not seeing many polls, and so due to recency being factored into the model, a newer poll in a polling drought will have a bit more of an effect than you might expect. (someone please poll this race!)
As a reminder, polls are broken into three parts: topline (50%), electorate screen (25%), and crosstabs (25%). We combine this with 2020 margins and education-based turnout fundamentals (calculated via legislative district turnout regressions) to get our projection.
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

Like this author's thread?

Get emails when @lxeagle17 has new unrolls!

Check out other threads from @lxeagle17!

Read all threads by @lxeagle17

:(