Strikes me as a situation where it could well be introducing bias but a) it's directionally correct and b) warts and all, better than not doing it at the end of the day.

I'd have to hear more about those differential rates + then think through about the realistic effect sizes of such a bias.

That said, the most consistent effect across racial lines isn't POC generally, it's black voters in particular and that lines up well with over-reporting patterns by state.

New piece from me @chriswarshaw.bsky.social , @devincaughey.bsky.social , and Bernard Fraga.

Story: though we've entered a high turnout era, turnout gaps between racial groups have reemerged.

Nerdy story: Using publicly available data, we've generated new, reliable turnout estimates over time.

Not sure what will count as substantive here but I think the addition of the voter validation stuff is valuable.

It would suggest black turnout a) it didn't surpass white turnout in 08/12 (suggested by the results you linked to) and b) it's now ~indistinguishable from AAPI turnout (which is novel).

2024 brought high voter turnout – but a growing racial gap New data shows that while white turnout has spiked, turnout among Black and other non-white Americans has stagnated.

The story itself: goodauthority.org/news/2024-br...

More methodological goodness and data downloads for you nerds: democracyfund.org/idea/the-com...

We thought it possible to build estimates that did a better job.

The biggest methodological jumps are a) we incorporated voter validated survey data to adjust self-reported turnout rates and b) found that we could make sub-state adjustment to the CPS.

OK, now the nerdy part.

While other estimates exist out there, they all have issues: Not comparable over time, not based on open methodologies, based on private data, etc.

It's not 100% clear why this is happening but we do offer a few potential explanations.

What's more, these estimates are likely underselling the story about black voters.

Our most exhaustive estimates - which incorporate voter validation data but are only available from 2006 onward - suggest that black turnout was even lower. Now likely on par with AAPI voters.

Bonus chart - the turnout gap vs. whites.

Black voters - who essentially closed that gap in '08 + '12 - are back to 90s levels.

The gaps for Latino voters have remained steadily wide. While APPI voters are doing better than earlier periods, '24 still represented a decline.

We've entered a high turnout period - something that should be seen as a positive development.

However, the dynamics of this hasn't resulted in equal gains among all groups. Specifically, 2024 turnout levels resulted in turnout gaps between whites and other groups going up.

New piece from me @chriswarshaw.bsky.social , @devincaughey.bsky.social , and Bernard Fraga.

Story: though we've entered a high turnout era, turnout gaps between racial groups have reemerged.

Nerdy story: Using publicly available data, we've generated new, reliable turnout estimates over time.

2024 brought high voter turnout – but a growing racial gap New data shows that while white turnout has spiked, turnout among Black and other non-white Americans has stagnated.

In recent years, white turnout has spiked. But turnout among Black and other non-white Americans has stagnated. @devincaughey.bsky.social, Bernard Fraga, @rpgriffin.bsky.social & I have a summary in @goodauth.bsky.social of our work on turnout in U.S. elections. goodauthority.org/news/2024-br...

Be cautious if...

-different polls are telling you different things
-results are really surprising
-results are about a ~hard to survey population

I'd say "Trump unfavorables dropped 20pts in two weeks with young voters" matches about 2.3/3.0 of those statements.

today.yougov.com/topics/polit...

Creates a lot of room for people to claim whatever they like while ignoring all the other stuff. Not ideal!

FWIW, I honestly have very little dog in any of these fights but it's frustrating to see the same dynamics over and over.

What I find frustrating about this particular topic is the number of "researcher degrees of freedom".

The party closer to median voters varies not only by topic (immigration vs healthcare) but even varies within topic like immigration (which is one of the foci of this piece).

As the power turns,
So do their concerns,
Economy's the best,
Or it's in distress,
Based on party lines.