The numbers in Kentucky absolutely do add up.

- Kentucky is a heavily Republican state (Trump +26 in '20)
- Eastern Kentucky has gotten increasingly more Republican
- Senate results have become increasingly correlated with presidential results

One of the points in the article is that McGrath underperformed party registration statistics in rural Breathitt County, KY.

But the explanation is simple: over last 20 years, many former Democrats started voting Republican, but they just didn't update their party registration. Image
Another example raised is Elliott County, which never voted for McConnell before 2020.

But this is also easily explainable. The county used to be Democratic, but has gotten more Republican because of: (1) Policy issues like coal (2) Trump's gains with white working class voters Image
This can all basically be boiled down to a few trends:
- Republicans have gained with white working class voters recently, starting at the presidential level.
- Voters have increasingly been voting for the party, not the person.

In places like Kentucky, this advantages the GOP.

• • •

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

Keep Current with Ryan Matsumoto

Ryan Matsumoto 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 @ryanmatsumoto1

3 Jan
I would say that the strongest argument in favor of Republicans is that Perdue won the first round by 1.78 points.

The pro-D counterargument is that early voting turnout data is strong for Democrats.

(1/6)

A lot of this comes down to how revealing you think the early vote data is.

On one hand, many people have been burned by EV analysis in the past. The GOP can just show up on Election Day.

But there are a few factors that make EV potentially more interesting this time...

(2/6)
First, we have a decent baseline comparison of the November election. It's recent enough that there probably aren't a ton of people changing their minds.

This makes Early Vote analysis better than comparing 2012 to 2016, for example, when many people switched sides.

(3/6)
Read 6 tweets
30 Dec 20
New Georgia Senate poll from @WinWithJMC:

Ossoff 53%
Perdue 45%

Warnock 54%
Loeffler 45%

This Georgia poll also asked how/when voters plan to vote.

Just 7% said they planned to vote on Election Day.

If that actually pans out, tough to see how Republicans win given the pro-Democratic Early Vote data.

But 7% just seems REALLY low.

20% of the votes in November came on Election Day. So 7% in the runoff would be EXTREMELY low.

The crosstabs have Perdue winning the Election Day vote 84% - 16% and Loeffler winning it 78% - 22%.

If way more than 7% of votes come on Election Day, that would benefit Republicans.
Read 5 tweets
24 Dec 20
The explanation for 18 of 19 'bellwether counties' breaking their streak of voting for the winner is pretty simple.

Compared to the country as a whole, bellwether counties are substantially whiter (82% vs. 60%) and less college educated (22% vs 32%).

docs.google.com/spreadsheets/d…
In the past, white working class voters voted more like the country as a whole, making these counties a good barometer for the country.

But in 2016, Trump made massive gains with white voters without a college degree, helping him win many bellwether counties by double-digits.
The bellwether counties actually became unrepresentative of America in 2016, when they voted 16 points to the right of the nation (versus 1 point to the left of the nation in 2012).

They only retained their bellwether status that year because Trump narrowly won the election. Image
Read 4 tweets
21 Dec 20
Georgia Senate Early Votes by Race (12/21 update):

White: 54.4% (56.5% in general election)
Black: 32.4% (27.7% in general)
Hispanic: 1.9% (2.6% in general)
Asian: 2.1% (2.6% in general)
Native American: 0.1% (0.1% in general)
Other/Multiple/Unknown: 8.4% (10.5% in general)
Democrats will be happy that Black turnout is relatively strong so far. But they'll want to increase Hispanic/Asian turnout as well.

But important to keep in mind that it's still VERY early - there are 2 more weeks of voting left and a lot can change!

It is important to be **very careful** when looking at early vote data. 80% of Georgia's general election votes were cast early/by mail, but the 20% cast on election day was heavily Republican. And the 80-20 split may be different for the runoff.

Read 5 tweets
10 Nov 20
Broke: Stories about Obama/Trump voters in Rust Belt diners

Woke: Stories about Clinton 2016->Trump 2020 voters in pho restaurants and taquerias
Trump's inroads with Latino voters, especially in the Rio Grande Valley in Texas, have been well documented.

It appears that Trump made some pretty significant gains with Vietnamese voters as well.

Here's a good article by @terrygtnguyen about pro-Trump Vietnamese voters and the complexities of the community's politics.

vox.com/first-person/2…
Read 4 tweets
6 Nov 20
So I guess the question is if they are able to keep scanning batches of 300 votes whenever they want, or if they have to wait 20 minutes each time.

#ElectionNight

Some Clayton County math:

If we have to wait 20 minutes per scan and Biden wins each 300 vote scan ~90% - 10%, he would net ~240 votes per scan, and we'd need 6 scans for him to cut Trump's 1267 vote lead down to zero.

That would be about two hours.

Say we're hoping to save 20 minutes. What would it take for Biden to overtake Trump in the span of five 300-vote scans?

1267/5 = 253.4
300/2 + 253.4/2 = 276.7
276.7/300 = 92.23%

So Biden would need to win ~92.22 to 7.33% to overtake Trump in five scans.

Read 6 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!