Nate Cohn

Nate Cohn @Nate_Cohn

Obviously there's a blurry line on all of these effects tests. Ok not 1%, but how about 2%? 5%? 10%? What about in cases where there's less obvious data on the magnitude of the effect? Future cases will test the limits, but by rejecting disparate impact they've opened the door — PolitiTweet.org

Nate Cohn @Nate_Cohn

Same story on whether there's a disparity on the effect by race. Some disparity is inevitable with any regulation, the court claims, so the size of the disparity may help inform whether it's bad. It rejects disparate impact as a test, alone. — PolitiTweet.org

Nate Cohn @Nate_Cohn

1) The court cares about the size of the effect. It says voters should deal with the 'usual burdens of voting,' and a big effect on turnout appears to be one way to evaluate whether it's an unusual burden; Or in short, if 99% of people get to the right precinct, nbd if 1% don't — PolitiTweet.org

Nate Cohn @Nate_Cohn

Some believed SCOTUS might take this opportunity to eviscerate Section 2. That didn't happen. But the courts reasoning offers some lessons about what kinds of restrictions on the time, place and manner of elections may or may not be legal going forward — PolitiTweet.org

Nate Cohn @Nate_Cohn

The Supreme Court upheld two Arizona voter suppression-type laws that Democrats sought to challenge under Section 2 of the VRA. Some initial, rough thoughts https://t.co/cp1LlJKoA2 — PolitiTweet.org

Nate Cohn @Nate_Cohn

The major explanation is educational polarization; Clinton actually did better among white voters than Obama! And educational polarization among whites appears to be continuing, at least at the geographic level (Pew actually dissents at individual level) https://t.co/uez1ja7cwS — PolitiTweet.org

Nate Cohn @Nate_Cohn

One could just as easily tell the kind of David Shor-ish story, where Democrats moving to the left led some number of relatively conservative, law-and-order sympathizing Democrats to defect to Trump (though I continue to think there's basically no good evidence on any of this) — PolitiTweet.org

Nate Cohn @Nate_Cohn

Now perhaps a more progressive Democrat would have maintained the loyalty of those voters. But it isn't obvious that an ideological appeal is the way to win the kinds of voters who flip to lean Republican and vote Trump, either — PolitiTweet.org

Nate Cohn @Nate_Cohn

The implication seems to be that Biden's losses among Democratic constituencies mainly involves voters shifting their partisanship from Dem/lean Dem to Rep/lean Rep, not merely Biden failing to win voters who still lean Dem. The reverse would be true of GOP constituencies. — PolitiTweet.org

Nate Cohn @Nate_Cohn

The partisan composition of the electorate (including leaners) is essentially unchanged, declining from D51-R48 to D50-R48. — PolitiTweet.org

Nate Cohn @Nate_Cohn

According to the Pew data, Biden actually gained among liberal and moderate Democrats, even while losing ground among Democratic *constituencies* like Latino voters https://t.co/vxhOYSzCNY — PolitiTweet.org

Nate Cohn @Nate_Cohn

In my replies, a lot of people seem to conflate losing ground among Democratic constituencies with losing ground among progressives or losing due to moderation, and it's not quite that simple https://t.co/7jc9q3hMTF — PolitiTweet.org

Nate Cohn @Nate_Cohn

Ultimately, we're never going to have a perfect picture of this election. But at this point, we do have most of the major sources now in and it'll be a while to really sort through what makes the most sense. — PolitiTweet.org

Nate Cohn @Nate_Cohn

And while Pew may have the best sampling method of all the various election studies, it may have the weakest ability to adjust for bias on the backend--others are adjusting at the precinct or substate level, which at least helps a bit with some stuff. — PolitiTweet.org

Nate Cohn @Nate_Cohn

But a lot of this stuff is harder to corroborate than other matters--you can't look at the county data and learn about gender, in the same way we can for Latinos. And this stuff is all still built on polling, which is plainly ... subject to a number of biases. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Perhaps the big question is... how right is this? There's a lot of reason to like the Pew data. And there's a surprising amount of corroboration for the basic contours of the story, whether from the exits, votecast, or Catalist, even though all of the various sources have issues. — PolitiTweet.org

Nate Cohn @Nate_Cohn

All considered, it's the most pronounced version of the conventional narrative we've seen yet: Biden struggling to gain ground among core Democratic constituencies, but making up for it with big gains among GOP groups — PolitiTweet.org

Nate Cohn @Nate_Cohn

According to Pew, 73% of Trump supporters turned out v. 68% of Biden supporters, based on vote history data. That turnout pattern is about the same as it was in 2016, if not slightly more favorable to Trump, per their data — PolitiTweet.org

Nate Cohn @Nate_Cohn

At the same time, the study shows Biden losing ground among nearly every major Democratic constituency, including Biden winning just 59% of Latinos, the lowest of any of the post-election studies — PolitiTweet.org

Nate Cohn @Nate_Cohn

There's also a big gender story here. The Pew data has the gender gap falling in half, and while that's an above-average version of the trend there's now a lot of evidence showing at least some version of that story — PolitiTweet.org

Nate Cohn @Nate_Cohn

In the Pew data, many of these Biden gains rival or exceed his far more prominent losses among Hispanics. That may reflect the biases of county-level data, which makes it pretty easy to notice Hispanic shifts (Rio Grande!) but far harder to notice shifts among other groups — PolitiTweet.org

Nate Cohn @Nate_Cohn

The Pew study shows Biden made large--even double-digit--gains among relatively moderate to conservative voting blocs, like veteran households, white Catholics, married men, and white men without a degree. — PolitiTweet.org

Nate Cohn @Nate_Cohn

New Pew data--one of the last of the authoritative datasets on the 2020 election--doesn't dispute the traditional narrative of the election, but instead shows the most pronounced version of it yet https://t.co/CBPIwlgeVt — PolitiTweet.org

Nate Cohn @Nate_Cohn

@rp_griffin right, though, to be fair, if there's no geograhpic error in suburbs/rural areas (and i'm not saying that's true, to be clear), then there's less reason to assume it's bad polling — PolitiTweet.org

Nate Cohn @Nate_Cohn

@rp_griffin one reason why i'm withholding judgment on this particular issue: their data shows trump gaining 11 points among rural whites, but losing 12 points with white suburbanites, so i'm not entirely sure these findings are inconsistent with the ecological evidence — PolitiTweet.org

Nate Cohn @Nate_Cohn

@Redistrict i haven't looked at this at all, but is this missing 'election day' vote distributed sporadically across the city or is it broadly across the bronx (in which case, i'd wonder whether they released their early vote) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@davidshor @dlwwatt based on my recollection from SF, the people voting paper boy are particularly unlikely to exhaust their ballots on first choice — PolitiTweet.org

Nate Cohn @Nate_Cohn

@pettigrew_stats though you'd expect his raw vote margin to say the same, and ultimately that's the only thing that matters in terms of whether he can be overtaken in the absentee vote — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @Taniel: This wouldn't negate the huge transfers Garcia got from Yang over Wiley, & from Wiley over Adams, as those transfers were comin… — PolitiTweet.org

Nate Cohn @Nate_Cohn

Neither does the NYC Board of Elections https://t.co/q0hLsvPX9g — PolitiTweet.org