Not much "racial realignment" in these new Pew numbers. And the youngest voters are not any more Republican despite weak support for Biden in the early trial heats.

The story is more complex than these graphs suggest. Let's take Black partisanship and do 3 things:

1) Crucially, separate ANES and CES surveys

2) Calculate the % Democratic within and without independents who lean Democratic

3) No smoothing. Let's see the raw data.

https://twitter.com/jburnmurdoch/status/1767198799934427371

Here's the graph. First, the "gold standard" ANES shows no secular decline in black partisanship. All of the decline is in the CES.

Moreover, the % Democratic is routinely higher in the ANES than the CES, especially if you include leaners.

In advance of election night, I think it's useful to calibrate expectations to the fundamentals, not to polls or vibes.

So, a good calibration:

House: Dems lose ~40-45 seats
Senate: Dems lose 1-3 seats

Anything significantly less = Dem over-performance. Here's why. A standard model for House elections is Gary Jacobson's: presidential approval, income trends, and # seats controlled by the president's party.

Forecast: Democrats lose 45 seats.

dropbox.com/s/95s7hip8bzg5…

Here's what I think would move the popularism debate along. We need some specific testable claims, because I see different versions of the idea circulating.

1) The popularity of the president's public positions is associated with their approval rating and reelection.

https://twitter.com/mattyglesias/status/1509512012589588483

2) The popularity of positions associated with a political party (fairly or not) affect the popularity of all of a party's officeholders.

(e.g., this is the claim implicit in the idea that "defunding the police" cost Democrats in 2020)

From a survey of Republicans conducted in January 2020, about one year ago:

51% agree that "The traditional American way of life is disappearing so fast that we may have to use force to save it." 47% agree that "Strong leaders sometimes have to bend the rules in order to get things done."

This is an interesting notion by @jbouie, but some challenges for "the stimulus helped Trump" theory are:

nytimes.com/2020/11/18/opi… (1) Based on the historical relationship between the change in disposable income and presidential election outcomes, Trump vastly *under*-performed.

New paper from @vavreck, @cwarshaw, and me: "The Effect of TV Advertising in US Elections."

Here is a brief thread.

chriswarshaw.com/papers/adverti… We estimate the effects of ads in:

5 presidential elections
276 U.S. Senate elections
176 gubernatorial elections
1,655 U.S. House elections
and 157 other state-level elections

All from 2000-2016.

I find the framing in this NY Times piece very odd. The notion is that GOP support for Trump's SCOTUS nominee reflects his "iron grip" on the party.

But, if anything, it's the opposite: the GOP's grip on Trump.

nytimes.com/2020/09/22/us/… Trump's ascendancy in the GOP depended on changing himself. He launched attacks on Obama, spoke on conservative media and CPAC, donated more to GOP candidates, endorsed Romney in 2012, etc. From Identity Crisis:

Inspired by some recent pieces by @amyewalter and @jimtankersley, I dug into Donald Trump's economic approval numbers, which are typically higher than his overall approval numbers.

Here's what I found: washingtonpost.com/politics/2020/… About 10% of Americans approve of Trump's handling of the economy but don't approve of Trump overall.

The challenge for Trump? They favor Biden by 41 points!

Pleased to announce the publication of a new @DemocracyFund research brief on public attitudes toward COVID, co-authored by my @VandyPoliSci colleague Cindy Kam and me.

Here's a link: voterstudygroup.org/publication/sy…

And now a thread on what we've found. We argue that public opinion about disease outbreaks can be understood as the intersection of biology and politics.

The story starts with Cindy's earlier research on public attitudes toward both the Ebola and Zika outbreaks: journals.uchicago.edu/doi/abs/10.108…

THREAD: This @MorningConsult survey experiment deserves more attention.

There are the expected partisan differences in support for protesting injustice.

But support drops -- especially among Republicans -- when black Americans are doing the protesting.

morningconsult.com/2020/06/01/flo… It reminded me of a related experiment we described in Identity Crisis. Clinton voters and Trump voters don't differ much on whether "average Americans" have gotten less than they deserve.

They disagree on whether African Americans have gotten less than they deserve.

Asking people whether something will affect their vote -- a campaign event, an issue, whatever -- is basically worthless. We're not good at reporting the reasons for our choices or whether something will change our mind.

And, yes, I am subtweeting a poll about Dem VP picks. Some relevant scholarly literature on why we shouldn't ask people this type of survey question. First, Nisbett and Wilson's "Telling More than We Can Know" (psycnet.apa.org/doiLanding?doi…)

Biden has been leading Trump among seniors since at least JULY 2019. So it's not just about COVID. New from @rp_griffin and me for @monkeycageblog:

washingtonpost.com/politics/2020/…

A brief thread. Here's the key graph, showing the trends in a Trump-Biden match-up over time. Younger voters have been moving away from Biden, but that's just bringing them to where they were in 2016.

For journalists writing about the new Trump "Death Star" ad blitz, it's imperative to discuss the research showing how quickly the effects of presidential general election advertising wear off.

The impact of advertising in May on what happens in November is likely zero. You can see this paper by @vavreck, @seth_j_hill, et al. (…2-4c2c-85dc-8326dbe5039a.filesusr.com/ugd/74791f_c5f…) our our books on 2012 and 2016.

Do you know what are very popular? State and local restrictions to enforce social distancing and reduce coronavirus infections.

A thread with data based on, you know, representative samples of Americans. The data are from Nationscape surveys (voterstudygroup.org/nationscape) from March 19-26 and April 2-11.

The question is "Some state and local governments have taken certain actions in response to the coronavirus and are considering other actions. Do you support or oppose the following"

Surveys are now clearly showing an increase in (1) unfavorable views of Asian people and (2) the belief that China is an "enemy." New at @PostEverything from @rp_griffin, Michael Tesler, and me. Here's a brief thread.

washingtonpost.com/outlook/2020/0… In the large samples (6000 per week) of Nationscape , the percent of Americans who rate Asians unfavorably is growing.

Parallel movement among Democrats and Republicans, at least until the last few weeks. (We don't see a similar shift in views of African Americans or Latinos.)

At @monkeycageblog, @rp_griffin and I dive into why Trump's approval bump isn't helping him much against Biden. washingtonpost.com/politics/2020/…

A story in three graphs, using the Nationscape data. First, views of the economy continue to plummet. They are now net negative even among Republicans.

New from me @PostEverything: Trump has long been characterized as the "people's billionaire" but today most Americans don't think he cares about "people like me." washingtonpost.com/outlook/2020/0…

A thread. Let's start with the data. The survey questions are how much a politician cares about the wealthy, the middle class, the poor, and "people like me." Let's compare views of Romney and Obama in Nov 2012 to views of Trump, Biden, and Sanders now.

New from @rp_griffin and me @monkeycageblog : "Americans have increasingly dismal views of the economy. But they don’t blame Trump." Here's a brief thread.

washingtonpost.com/politics/2020/… Using the Nationscape data, we chart a huge increase in economic pessimism in the last month -- including among Republicans.

In my Twitter feed, many people are saying anti-Trump ads like this Bloomberg bullying ad won't work because Hillary's anti-Trump ads didn't work in 2016.

That is not consistent with the evidence from 2016.

https://twitter.com/MikeBloomberg/status/1228755924115955712

In our book, @vavreck, Tesler, and I showed that HRC's ad buys at the media market level were positively correlated with her county-level vote share, controlling for Obama's vote share in '12, Obama and Romney's ad buys, and a variety of other factors.

amazon.com/Identity-Crisi…

On the day of the Iowa caucus, here is a graphical tweetstorm of the Democratic party's invisible primary.

In short, the party didn't really decide. To the extent that it did, Biden was more or less the winner.

Now, to the graphs! This is the pace of endorsements pre-Iowa by Dem governors and members of Congress. The invisible primary developed pretty slowly, and in the end relatively few leaders endorsed (30%). 2020 ends up looking like 2008.