Share this page!

G. Elliott Morris Profile picture
Twitter logo
11 Dec, 4 tweets, 2 min read
not an easy day to focus on book writing. is this how political scientists have felt for the last 4 years?
Not only is the statistical analysis that Paxton relies on incomprehensibly misspecified, but the author actually explains *why* it's wrong (later votes came from cities, competitive battleground metros are less latino) IN THE CONCLUSION OF HIS STATEMENTS

supremecourt.gov/DocketPDF/22/2…
another reminder not to let economists near voting data
this is a little like rolling a ball down a hill and saying it will continue to roll until the end of time, without taking into account that the hill eventually gives way to a flat meadow — and then when the ball DOES stop rolling, concluding that the law of gravity is fraudulent

• • •

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

Keep Current with G. Elliott Morris

G. Elliott Morris 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 @gelliottmorris

G. Elliott Morris Profile picture
28 Nov
The think-pieces about Why It Was #ACKSHUALLY Wrong To Freak Out About Trump Stealing The Election seem needlessly contrarian and exceedingly misguided. It is very, very worthwhile to actively worry about a low probability event if the result of it is the end of our democracy.
“Haha suckers, see? Trump did everything he could to totally upend the Constitution and it didn’t work! Only an entire political party believes him and the future of our democracy is clearly imperiled... But see? I told you not to worry!!”
Nearly half of voters and political leaders in America are complicit in an attempted coup of a presidential election — a scheme that may have worked in just a slightly closer contest. Not sure now is the time for cheerleading.
Read 4 tweets
G. Elliott Morris Profile picture
11 Nov
Our latest YouGov/The Economist poll has a host of troubling findings about public confidence in the election.

Most shocking is that 86% of Trump voters say that Biden "did not legitimately win the election." 73% say that we'll "never know the real outcome of this election."
We also see the usual patterns in attitudes about mail-in voting and fraud. 88% of Trump voters say they believe that "illegal immigrants voted fraudulently in 2016 and tried again in 2020," for ex, and 90% believed that "mail ballots are being manipulated to favor Joe Biden."
Republicans are also exhibiting some... concerning... attitudes about the franchise, with 46% saying that "some people are not smart enough to vote" (27% among Dems) and 43% saying that people should have to pass a test before voting (15% for Dems).
Read 4 tweets
G. Elliott Morris Profile picture
10 Nov
This map shows you which parts of the country swung to the left or to the right v 2016. Georgia and sub/urban Texas really stand out to me:

economist.com/graphic-detail… Image
Ohio and Iowa are also notable for their shift to the right, which is doubly striking since the polls said they had swung so far to the left
(we are predicting results in counties that haven't fully reported totals yet)
Read 6 tweets
G. Elliott Morris Profile picture
10 Nov
It is weird that so many people who rely on public polling for their jobs are instantly willing to trash the industry with damaging statements like these, instead of trying to calmly and transparently explain what went wrong and work toward better polls in the future.
Like, if part of the answer to "why aren't people taking surveys?" is "because people in the media trash them and say they are biased and aren't good for anything," then it's weird that people who know better would actively participate in the industry's demise
For posterity's sake,

(1) Here is our first cut at how the polls/models missed support for Trump, again: economist.com/united-states/…

(2) Here is what we're going to do differently next time: economist.com/united-states/…
Read 5 tweets
G. Elliott Morris Profile picture
7 Nov
It’s way too early for headlines like this. Polls are going to end up with only slightly-above-average error once all the votes are counted. The emerging pattern of routine bias is a problem, yes, but it’s not the doomsday situation people are leading you to believe.
The biggest problem with polls TBH is that people, and lots of journalists, don’t understand how much error there can be. This is a perfectly reasonable performance for the data, yet so many in the media have already prematurely called the industry dead.
The other point is re: the “who” of the error. Based on county returns so far it looks like polls actually performed just fine in the suburbs, but underestimated support for Trump among non-college whites and maybe Hispanics. Nothing about revolt against D culture looks true.
Read 5 tweets
G. Elliott Morris Profile picture
5 Nov
The primary use for polls is not pre-election handicapping, it's to understand what the average American wants from their government. And while election polls aren't in a good position, issue polls of all adults seem fine (for now).

bloomberg.com/opinion/articl…
I agree that people need to "kick the addiction" of thinking polls are infallible predictors of election outcomes. They're not, that's why we try to model what could happen if they misfire.
Separately, I have noticed this really annoying trend where a lot of the people saying we should "quit" polling are finance or quant folks who want 100% accurate predictions from them, but that's not what polls are and never will be.
Read 5 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 @gelliottmorris has new unrolls!

Check out other threads from @gelliottmorris!

Read all threads by @gelliottmorris