Nate Cohn

Nate Cohn @Nate_Cohn

@RyanDEnos @blfraga (that's among all RVs; 24% of voting 'unknown' are classified as black) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RyanDEnos @blfraga i did/have; it's not materially different. the unknowns are 28% black per L2's classification (and our poll responses even less so) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RyanDEnos @blfraga yes--the chart in the original article just uses the GA SOS data without imputation, though i do 'use' the L2 data for analysis — PolitiTweet.org

Nate Cohn @Nate_Cohn

@zachacrowell @jon_m_rob @blfraga i think the difference on ours is < .5, so i'm not sure there's enough precision to know if there's a difference here. the turnout rate by race does seem consistent with a decline in black share with a more precise account, tho idk their denoms — PolitiTweet.org

Nate Cohn @Nate_Cohn

We have three kinds of data: --our survey responses (we call off the file, so we can compare voter file v. poll race) --the commercial modeling by L2, our voter file vendor --our own modeling, mainly based on the black share of an unknown voters' precinct https://t.co/qfTpmOnZto — PolitiTweet.org

Nate Cohn @Nate_Cohn

Ultimately, dwelling on this very legitimate question winds up obscuring from the fact that the conclusions aren't contested. It let's you say you think the analysis is 'wrong,' without being willing to say that the black share of the electorate didn't drop — PolitiTweet.org

Nate Cohn @Nate_Cohn

Ultimately, dwelling on this very legitimate question winds up obscuring that the conclusions are all but uncontested. It let's you say you think the analysis is 'wrong,' without being willing to say the conclusion is wrong--and in this case, actually stipulating it's right — PolitiTweet.org

Nate Cohn @Nate_Cohn

Ultimately, dwelling on this very legitimate question winds up obscuring from an all but uncontested conclusions. It let's you say you think the analysis is 'wrong,' without being willing to say the conclusion is wrong--and in this case, actually stipulating it's right — PolitiTweet.org

Nate Cohn @Nate_Cohn

But I don't want to debate about whether the 'unknown' vote is 26 or 35% black, or whatever, because a) it's not knowable (they're unknown!); b) it doesn't make a difference: all of the hard, factual ways of analyzing the electorate show a dip in the Black share of voters — PolitiTweet.org

Nate Cohn @Nate_Cohn

Well, as I said, if you go down the road of allocating the 'unknown' voters you wind up having to debate about it. The data at my disposal says they're no more likely to be black--and probably less--than voters of 'known' race. https://t.co/JJPeUpuKBP — PolitiTweet.org

Nate Cohn @Nate_Cohn

In the majority black wards of Milwaukee, turnout fell beneath 2016 levels. In Philadelphia's majority black wards, there was basically no change. TBH, the only interesting question is whether these spots are in more severe population declines than the census says — PolitiTweet.org

Nate Cohn @Nate_Cohn

This isn't specific to Georgia. It appears to be true in NC, where the data is of similar quality and nearly all of it is in. It's probably true in MI/WI/PA, where the data isn't of similar quality but the same geographic pattern--relatively weak turnout in MIL/PHI/DET holds — PolitiTweet.org

Nate Cohn @Nate_Cohn

At the end of the day, I don't see any actual disagreement over any of the words in the article or its conclusion: the black share of the electorate dipped, to its lowest level in a while, bc black turnout didn't surge to the same extent as nonblack turnout — PolitiTweet.org

Nate Cohn @Nate_Cohn

As long as I've been at the Times, none of my GA analyses have allocated 'unknown' voters. That's in part because you wade into the debate about how to do it, which seems unnecessary if it doesn't materially affect your findings--as in this case — PolitiTweet.org

Nate Cohn @Nate_Cohn

To the extent there's a disagreement (and I wouldn't even call it that), it's about a very valid question about when (and potentially how) to allocate voters of 'unknown' race in GA. Whether you do so doesn't affect the conclusion, but it's being used to imply that it does — PolitiTweet.org

Nate Cohn @Nate_Cohn

No actual disagreement https://t.co/GOKn3c6Efx — PolitiTweet.org

Nate Cohn @Nate_Cohn

When it's all done, it'll be interesting to estimate how large the E.C.-P.V. gap would have been if Trump had maintained '16 levels of support among Latino voters. It's not crazy to think it might have reached 5 pts, esp since Trump made relatively limited gains among AZ Latinos — PolitiTweet.org

Nate Cohn @Nate_Cohn

When it's all done, it'll be interesting to estimate how large the E.C.-P.V. gap would have been if Trump had maintained '16 levels of support among Latino voters. Doesn't seem crazy to think it would have been a 5 point gap — PolitiTweet.org

Nate Cohn @Nate_Cohn

It is a surprise in one respect: Trump had such a large, relative edge in the Electoral College despite making huge gains among Hispanic voters, who are generally concentrated in non-competitive states — PolitiTweet.org

Nate Cohn @Nate_Cohn

In the final account, Trump will fare around 4 pts better in the tipping point state (Wisconsin) than he did nationwide. It's a huge gap, but it's not necessarily a huge surprise — PolitiTweet.org

Nate Cohn @Nate_Cohn

With all of that cause for humility in mind, the assumptions underpinning the opinion piece cited at the outset are... really going beyond what we can do with the data. And it's pretty unlikely given what we do know — PolitiTweet.org

Nate Cohn @Nate_Cohn

Unfortunately, I don't think we're going to resolve these questions to the extent we could in prior elections, even when all of the data arrives, and that's even before considering the extent that the modest changes from 2016 raised the level of precision we needed to do well — PolitiTweet.org

Nate Cohn @Nate_Cohn

I think we have every reason to assume they were off by a lot on older voters. But we don't have proof of that, and we don't know by how much. — PolitiTweet.org

Nate Cohn @Nate_Cohn

But we that's only what we know about, based on the limited, exact data I mentioned earlier. We have no idea whether they were also off by a lot on say, young voters or women. — PolitiTweet.org

Nate Cohn @Nate_Cohn

We know the national surveys were off by 4 points or so. Worse, we have reason to believe they're off in very specific ways--like a severe underestimate of Trump's strength among white working class voters — PolitiTweet.org

Nate Cohn @Nate_Cohn

Usually, most of the exacting analysis is done with pre-election survey data! Unfortunately, this year's pre-election survey data is as bad as it's been in a competitive election in a very, very long time--and in ways that are quite material for unpacking the story — PolitiTweet.org

Nate Cohn @Nate_Cohn

Unfortunately, this is a terrible, terrible year for survey data. The exit polls have always been bad and inappropriate for exacting analysis. This year they're both worse and incomparable to earlier exit poll data, rendering them.. even more useless — PolitiTweet.org

Nate Cohn @Nate_Cohn

We really have no hard, exact data that tells us about shifts among these groups. To figure this stuff out--and tbh, to increase our confidence in some of our geographic/ecological inferences earlier--we have to rely on survey data. — PolitiTweet.org

Nate Cohn @Nate_Cohn

But there are actually a lot of things we can't say with this data. Age and gender are two great examples. They're just not very strongly correlated with geography. — PolitiTweet.org

Nate Cohn @Nate_Cohn

With this data, we can see that Biden gained in well-educated suburbs, and especially traditionally GOP ones like ATL/KS/DAL/IND. We can say that Trump did well in Latino areas. We can say black turnout growth lagged non-black turnout surge, etc. — PolitiTweet.org