So over the past couple of days, Nate Silver has been tweeting multiple times a day insisting that things are looking worse and worse for the Harris campaign.

One state he claims has moved 1.2 pts to Trump since the convention is WI. Let's take a look at the recent polls there. 6 of 10 post convention polls show Harris leading, by an average margin of 7 points. But the two GOP highschool kids' poll has the race even, Trafalgar (leaders of the red wave polls of '22) have Trump up 1, and then one Dem pollster I'm not familiar with had the race even.

We now have voter file updates for 38 states that cover the week of July 21st, affording us the opportunity to quantify the reaction of the electorate to VP Harris as the nominee. Our analysis compares new registrations during the week of 7/21/2024 to the same week in 2020, for the sake of consistency. Overall, total registrations were actually slightly down in 2024, relative to 2020 (289k to 283k). But what do the subgroups show?

At long last, we have a PA voter file update that includes the week of July 21st, after VP Harris ascended to the top of the ticket.... Okay, first, when compared to the same week in 2020, total new registrations increased by 34.3% in PA in the week following Biden's withdrawal. Dem registration increased by 46.6%, GOP by 21.1%.

The Harris Effect - in the 13 states that have updated voter files since July 21st, we are seeing incredible surges in voter registration relative to the same time period in 2024, driven by women, voters of color, and young voters. Young Black women are leading the way, seeing their registration almost triple, relative to the same point in 2020. Young Hispanic women aren't far behind, with a 150% increase in registration. Black women overall have almost doubled their registration numbers from 2020.

Okay, when I have time I'll do a deeper dive on post-candidate changeover voter registration trends, because it seems we are witnessing something important happening, and while we obsess over polls, quantifying actions taken by voters can be more helpful.

First up, Michigan. Looking at the week after July 21st, we saw an immediate increase in women registering to vote. In that time period, the gender gap was +7 women, while there was effectively no gender gap in new registrations during that same week in 2020. We also saw 26% more new registrations.

Wow. We have a decent number of states with recent registration file updates now. I ran the gender gaps among new registrants in the two weeks after Biden's withdrawal. Women have been registering to vote at a very high rate. This is eclipsing the post-Dobbs surges most places. These are the states that have released post July 21 updates so far. I will continue to update this with new states as the updates come in.

Are you ready to play the "let's overreact to very small samples of data" game? I have been impatiently waiting for updated new voter registration data from post July 21st to gauge the electoral reaction to VP Harris as the nominee. But it takes time for states to update... I'll add states to this thread as they come in. First on the list... Alaska! In the week after July 21 in 2020, new registrants were +6 GOP. In that same week this year, they were +7 Dem. In '20 it was +2.5% women, this year it was a shocking +12% women.

We finally have the individual turnout data from the NY3 special election (where Dem Tom Suozzi won in George Santos's open seat). I found a couple of data points very interesting. First, let's remember that Dems ran heavily on abortion in this seat, which Santos won in 2022 thanks to very low Dem turnout.

On the 2nd anniversary of the Dobbs decision, I figured I'd share a reminder of how abortion rights has become the most powerful single issue in politics, and is likely more salient now than it was in 2022. When the decision was handed down two years ago today, it was a shock but not a surprise, given the leak several weeks earlier. We were left with an open question of how it might impact the 2022 elections. The first answer would come 39 days later, from Kansas of all places.

Debunking the popular theory that Nikki Haley's relatively substantial support in GOP primaries has been due to Dem leaning voters flooding the GOP primary vote, thereby suggesting that Haley's strong showing isn't actually a bad sign for Trump... We recently analyzed the GOP primary individual turnout history in the NC primary. Haley won 23% there. Looking at the unaffiliated voters who comprised one-third of the GOP vote, there is no indication that these voters were Dems seeking to stop Trump.

Here's my grand* unified theory of this presidential race, and why I feel like the Biden team should feel pretty good about their standing:

*this is not all that novel or deep Polls are consistently showing Biden faring much better among "likely voters" relative to all registered voters or American adults. We can quibble about what a likely voter is, but the same thing is apparent if you filter on past turnout history.

I've got a few mins, so I'll share my observations as I go through the new Suffolk poll released this morning.

suffolk.edu/-/media/suffol… On the first page of tabs, a look at the sample. Remember, this is registered voters. So voters age 65+ are 22% of this poll's sample, but will be closer to 35% of ballots cast. More in a second on what that means...

Lots of fun dismissive jokes today about "crosstab truthing" towards anyone raising any concerns about this poll, but this is perhaps the more insidious counterweight to questioning crosstabs: lending them lots of credibility when they support your priors, ignore when they don't.

https://twitter.com/admcrlsn/status/1764096001575383525

If you're going to write a story about how Biden is hurting with young voters based on one poll, shouldn't you write a follow up story now talking about how Biden has miraculously rallied back support among young voters? (trick question, you should write neither)

Some of the loudest messaging right now is Dems/progs demanding that we all admit that Biden is old, and "deal with it".

Here's a newsflash - everyone is aware of his age. I understand why it might worry you, because the stakes are so high. But what are we asking for here?

https://twitter.com/tribelaw/status/1759221750095618325

Not a single voter will be convinced to change their mind if they think Biden's age is a concern. They can be convinced of two things:

- All evidence suggests he has been a very capable president, regardless

- The alternative is far, far worse

So maybe let's focus on that?

NY3 Special Election Eve Wrap Up Thread:

The early in person votes have all been cast now, with >60k votes ready to be counted. By party reg they are +10 Dem (ie, the share of early votes cast by reg Dems is 10 pts higher than that of GOPs). That's a 2.8 pt improvement on '22.

https://twitter.com/tbonier/status/1756849693970006059

The mail votes will continue to trickle in over the coming days, but I wouldn't expect to see more than a few thousand more come in. About 13k have been returned, and by party reg they are +29D, compared to +26 in '22.

Perhaps unsurprising to anyone, I have lots of issues with Matt Yglesias' latest piece "Joe Biden is Currently Losing the Election". Some disagreements based on perspective, some on outright errors. I will agree with Matt on one broad point - we should all be operating with the understanding that Trump can win this election, and it won't take some far-fetched scenario for that to happen, and act accordingly.

It was nice of Taylor Swift to engage in a conspiracy with the NFL, the Biden administration, and Pfizer, to fix the Superbowl in KC's favor, in order to make up for the regular season games that she was responsible for them losing. (Did I do this right, MAGAs?) Also, Fauci.

I'm seeing more and more voices on the left suggesting that the "smart" thing to do is to abandon Biden and let Trump get elected, to show the Democratic Party that it needs to better represent their interests. Beyond the inherent danger in this strategy, it's also deeply flawed. If you walk away from the table, those who remain make the decisions. The likelihood that Dems will move to the left if Biden loses is not high. Nader acting as a spoiler for Gore in '00 led to John Kerry as the nominee in '04.

The KY turnout history for the '23 statewide election is posted! You'll recall this was the election where Dem incumbent Gov Andy Beshear won by a 5 pt margin.

Some quick takeaways from the turnout data... Dem turnout was strong. The 2023 electorate, based on partisan registration, was +2.9 GOP. While overall turnout was almost identical to '22, the '22 electorate was +6.8 GOP.

This knocks down one the immediate post-election hot take that Dems did well despite weak turnout.

I've been thinking about this poll data for a few days, and want to share some data to explain why I am skeptical of the phenomena we are seeing that shows Trump performing well with low turnout propensity young voters. Bear with me..

https://twitter.com/Nate_Cohn/status/1741507483456196751

First, let me make it clear that I'm not calling Nate out here or suggesting he's doing something wrong. I applaud his transparency. He's simply sharing the NYT polling data, which is interesting and illuminating.

Regarding the polls showing Trump winning as much as 31% of the Black vote, one of the things @SimonWDC and I have been saying for over a year now is that we have to pay attention to more than just the polls. Here's an example of what we're talking about... If there was truly such a radical realignment happening with Black voters, you would expect to see evidence, not just in polls. Yet, all of the other data sources show unwavering support for Dem candidates from Black voters.