Washington Post poll in Virginia is a bit of an odd one, and a departure from their state polling in prior cycles
Most obviously, they've gone from pure RDD in 2020 to an RDD+RBS sample, presumably to gain some of the benefits of the voter file while preserving full coverage.
Whatever the merits of that choice (most state telephone pollsters just go with RBS only), they try to blend the two samples in an odd way: they de-dupe the RDD frame, weight each to the full population, toss together (as oppose to voters with or without RBS coverage + p sltn)
I could go down a 50 tweet rabbit hole thinking through whether this design holds up as a theoretical proposition. Maybe I'd decide it did by the end of the rabbit hole. I'm not going to, though, because I don't think (?) there's any material consequence. It's just odd
The risk I wonder about: do you wind up reversing non-coverage bias and bias the sample toward the non-covered population, even if only slightly, if you dedup and weight the RBS sample to the full partisan file, and don't have a party weight on the non-RBS component
Re-reading the methodology, I misinterpreted the dedup initially—they’re still calling RDD respondents who have a number on the file as long as they’re not in the RBS frame. Reduces the risks—and benefit tbh—of the approach
• • •
Missing some Tweet in this thread? You can try to
force a refresh
A little early to tell, but if you squint at the numbers there's a case that Biden's numbers are inching back up a bit as Afghanistan fades from the news.
Here, Monmouth with Biden at 46/46 monmouth.edu/polling-instit…
Sure, there's probably some nonresponse bias. There always is!
But the amount of focus on partisan nonresponse bias has gotten way, way out of hand on this website as of late
Way back in the day, you needed a little bit of evidence to claim non-response bias was at play. Stability in the YouGov poll, or another party weighted poll, would offer some evidence--even if not conclusive evidence, given other methodological differences
With Election Day vote largely counted, "no" still has a considerable 64%, down from 67 at race call.
That may come down a bit--Biden fell about 1.5 pts between now and the outcome. OTOH, mail ballots over the last few days didn't seem as disproportionately GOP as 2020.
We'll see what it looks like in the end, of course, but at the moment I don't really see any signs that Newsom fared particularly poorly among Latino voters, as the exit poll suggested.
The results in relatively Latino counties appear about the same as 2018
And in the end, it may be that the polls underestimated 'no' by a fairly modest margin. Just typing that out, it's hard not to wonder whether acquiesce bias--the tendency for people to say 'yes' to be agreeable--might play a role. But there are many other plausible explanations
The polls have closed in California's gubernatorial recall election.
NYT results here >>> nytimes.com/interactive/20…
Expect Newsom to take an early lead, as the disproportionately Democratic early mail votes will be the first counted.
If the 2020 election is any indication, the 'no' vote could start out 5 pts higher than the ultimate result.
The exits show "no" up by around 18 points or so.
If so, those early Newsom votes may be enough to yield an early call cnn.com/election/2021/…
A flash-back to around 1AM ET on 11/4, with some modest implications for following the vote tonight
The early dump of advance votes was overwhelmingly for Biden.
In the end, Biden won California by a more modest margin--a mere 29 points--as the Election Day and late mail votes ate away at his lead, as it did in Arizona.
When you piece it all together, the Election Day vote was basically 50/50 and the late mail votes were more like 60/40. The early mail dumps were 70/30.
Looks like we're finally going to get to see the new Democratic voting rights compromise nytimes.com/live/2021/09/1…
In theory, it should put progressives more-or-less where they would have like to have been in February: unified behind a bill that Manchin could be a) enthusiastic about and b) convinced the GOP ought to support on the merits in a functioning chamber
If those two conditions are met, you theoretically open up that 1-in-10 chance--or whatever it may be--that Manchin could seriously consider reforming the filibuster to secure its passage
In a lot of ways, this article retreads a fairly well-established and uncontested history. But there is a twist here that I'd like to make explicit: it centers the rise of four-year college as the driving and explanatory force behind the rise of edu/cultural polarization
In a slight contrast, consider this very good piece from @edsall today, which tells the familiar tale of the rise of postmaterialist values without really talking about the role of rising educational attainment in shaping the content of postmaterialism nytimes.com/2021/09/08/opi…