Over the last month or so, the issue environment has kept shifting toward the Republicans. Whether that continues -- and whether it ultimately yields GOP polling gains -- is one of the biggest questions as we head into the final stretch nytimes.com/2022/09/27/ups…
There is pretty clearly a shift in the issue environment. Here, one example -- interest in abortion is back down to pre-Dobbs levels on Google Searches, with the economy overtaking it. It's quite a bit like the pre-Dobbs leak numbers
So far, there aren't huge signs of this reshaping the race. There are a few, perhaps -- Biden's ratings aren't going up anymore; Democrats losing ground in a few key Senate races v. August; maybe the RV/LV gaps in the polls this weekend. We'll have to see
A lot of you don't love the Google metric -- I like it a lot -- but here's another: the share of voters citing abortion as the 'most important' issue fell from 8% in July to 4% in September, per Gallup
• • •
Missing some Tweet in this thread? You can try to
force a refresh
If you want to see why the size of the bias in seats is relevant, just look at the The Economist model. Democrats currently at a 50.1 to 49.9 tie in the popular vote, favored to win 214, and the favorite *in* an outright majority of districts. Very fair
If the median vote share was moved to the national average (ie: shift MI08, OH13, NJ07 to d+4.5)... the Democrats still wouldn't be favored.
What would have 2x that effect? Moving 10 D+6 to D+4.5 CDs to D+2.2.
And very minor shifts in turnout in blue CDs /individual races would be sufficient to make the Democrats favored to win without the popular vote in this model
In today's newsletter, a nautically themed report on whether you can trust the polls -- including relatively good news on response rates by party in the latest Times/Siena survey nytimes.com/2022/09/20/ups…
In this poll, white Democrats were about 5 percent likelier to respond than Republicans.
That's quite a bit less than fall '20, when Democrats were 20 percent likelier to respond.
It's more like Oct '19, when Ds were 6 percent likelier -- and our results were better for Trump
Today we're launching a newsletter on elections, politics and polling as we head into the heart of the election season. I hope you'll sign up! nytimes.com/2022/09/12/ups…
I know there are lots of newsletters nowadays but I do think that the newsletter format is especially well-suited to political/polling analysis. It's easier to offer provisional and uncertain takes when you don't have to fit the stuffy, authoritative prose of a news article
It'll also be a natural home content that doesn't have a spot on the Times homepage--the sort of stuff that I've just tweeted out in the past:
--announcements about polls
--wonkier stuff, like the findings from our 2020 autopsy
We can also have some fun.
It's been exciting time over here in NYT elections land, where we've been getting our house back in order heading into the heart of the election season nytco.com/press/announci…
Believe it or not, we've never had full-time staff dedicated to our already extensive elections analytics operation. We managed to do hundreds of polls and dozens of needles anyway, and we're very proud of it, but the prior setup was neither desirable nor sustainable
So we've reorganized: we now have a formal Elections Analytics unit led on an interim basis by @williampd, who has been absolutely fantastic. He's contributing in ways never imagined, like figuring out if we could say "woke" in a poll in Spanish on the ground in Buenos Aires
I thought Democrats had a very good chance to win NY-19, so I'm not going to frame this as some massive upset.
But the reason it's not a massive upset is because the signs of Democratic strength just aren't rare anymore.
Democrats have overtaken Republicans on the generic ballot. They've been outperforming in several straight congressional races. The Washington primary results were decent. The Senate polls look almost-too-good for them.
As far as November is concerned, it's hard to weigh all of these kind of imperfect indicators against the long history of out-party strength in midterms. And yes, I think there are very good reasons to doubt some of these polls, especially in the Midwest.
The GOP holds MN-1 in last night's special election, but only by a modest 4 point margin (Trump+10 district; R+3 in last House race)
The signs of a Democratic rebound post-Dobbs are starting to pile up nytimes.com/interactive/20…
There haven't been many other special/non-primary election results since Dobbs, but MN-01 isn't exactly alone. NE-01 was also a strong showing for Democrats. There's also the KS abortion referendum, if you count it.
We'll get more data, including NY-19, over the next few weeks.
Democrats have also trended upward on the generic congressional ballot, where they've reached parity with the GOP
No way to know if it lasts until November, but the focus on abortion/Jan 6 hasn't ebbed--yet. At the same time, the news on inflation has improved for Ds