Lightnews — Scholar-powered news

Reposted by Will Stancil

Alex Schieferdecker @alexschief.bsky.social · 6h

Ghost of Christmas Past: "Look I'm going to be honest you guys, there's not a lot of material here to work with"

Will Stancil @whstancil.bsky.social · 6h

It was presumably also foolish to bet against Leni Riefenstahl for a while, but aligning yourself with the propaganda machine of the regime doesn’t take any great genius or insight, just a complete lack of scruples

1 11 130

Will Stancil @whstancil.bsky.social · 6h

Weiss (and Flanagan) spent years raising the alarm about a set of totally imagined threats to free speech. Then Weiss immediately aligned herself with incredibly powerful, incredibly corrupt men who actually stuffed free speech into a wood chipper, and they rewarded her with power. Brilliant!

Sally Jenkins @sallyjenx.bsky.social · 10h

Ratio me. Please. It's a badge of honor. If you don't like a link, go follow some chicken-heart who needs the approval of the thought-police. Caitlin Flanagan of @theatlantic.com is a tremendous writer and this piece is an excellent read. www.theatlantic.com/ideas/archiv...

Don’t Bet Against Bari Weiss

The new editor in chief of CBS News triumphs over her critics.

www.theatlantic.com

6 24 200

Will Stancil @whstancil.bsky.social · 6h

“Trump haunted by visions of hell” is my favorite Trump, although I’m not totally sure what’s going on with the dude

Aaron Rupar @atrupar.com · 6h

Trump: "I don't think there's anything that's gonna get me in heaven. I think I'm not maybe heaven bound. I may be in heaven right now as we fly on Air Force One. I'm not sure I'm gonna be able to make heaven."

99 160 1.3K

Reposted by Will Stancil

Matt Darling @besttrousers.bsky.social · 9h

Right - groceries alone was 5.86% of income in 2020 ($4,942/$84,352) and 5.95% of income in 2024 ($6,053
/$101,805).

7 4 66

Will Stancil @whstancil.bsky.social · 9h

To be clear this is spending on food, which partly increased because people were going out to eat more

4 27

Reposted by Will Stancil

Rising FRIGHTy Kuuga @kuuga.bsky.social · 10h

Because "all/most elections in 2024 went this way" may be true, but if you want to pick out a causative factor, you need to separate out variables, and that necessitates comparing to other elections through history and doing tons of material and cultural analysis, which is very difficult

1 17

Reposted by Will Stancil

Rising FRIGHTy Kuuga @kuuga.bsky.social · 10h

Now if you cast a dragnet back through history + across countries and you find a large number of other elections where first-time-inflation-sufferers voted against incumbents then we can talk turkey. But If your theory is only using 2024 as a data point then it's not a robust theory.

3 1 8

Reposted by Will Stancil

Rising FRIGHTy Kuuga @kuuga.bsky.social · 10h

When you narrow your statistical base to a small number of "eligible" data points you can make the stats say anything you want. When you say "a generation that has never experienced significant inflation" you have narrowed the data points to just one or two elections.

1 1 12

Will Stancil @whstancil.bsky.social · 10h

The whole “look at the 70s” thing is precisely why I’m skeptical. Because if there was some kind of clear, predictable historic trend here we could see it. But there isn’t

2 4

Will Stancil @whstancil.bsky.social · 10h

Josh people have modeled the effect of inflation on public sentiment AND THOSE MODELS STOPPED WORKING IN 2023-24

4 1 6

Reposted by Will Stancil

JB stan account @johnbrownstan.bsky.social · 10h

2024 was about inflation in exactly the same way 2016 was about email security

Will Stancil @whstancil.bsky.social · 13h

“Inflation caused 2024” is a comfortable and appealing explanation for a lot of people - it reinforces their patronizing sense that most voters are selfish little piggies with no inner life, it promises that things will be back to normal next time - but there’s really not any evidence of it

5 13 150

Will Stancil @whstancil.bsky.social · 10h

Yes it turns out science is hard and social science is much harder

4

Will Stancil @whstancil.bsky.social · 10h

More p-hacking

João Guassi Moreira @jfguassimoreira.bsky.social · 10h

I think I’ve seen work out there showing that even though the rate of inflation dropped from ‘22 - ‘24, voters felt prices were still high (which…they were because the rate slowed but prices didn’t drop), and the dem campaign didn’t a good job of communicating economic wins

3 1 15

Will Stancil @whstancil.bsky.social · 10h

Ultimately what you people are doing is just Vibe Analysis: taking two things that you feel certain should be related and asserting they are, without the barest hint of an actual useful model of how they interact, which could be used to make predictions or understand the world. It’s bad thinking!

5 1 34

Will Stancil @whstancil.bsky.social · 10h

“Inflation was why 2024 turned out the way it did.”

“Okay so what happens if inflation goes up or down in the future?”

“I don’t know.”

“Why doesn’t it match previous trends?”

“I don’t know.”

“Why didn’t inflation going down help Dems?”

“I don’t know.”

3 1 35

Will Stancil @whstancil.bsky.social · 10h

Josh my argument isn’t “absolute.” I’m saying that if inflation explains the 2024 result, or was some major causative factor, draw some conclusions from that, tell us what it predicts about the future, or how it matches past trends. Because otherwise you’re just waving your hands around

Josh Marshall @joshtpm.bsky.social · 10h

Not sure either. But these absolute arguments are ones that can't really convince any intelligent/engaged person. The arguments you summarize here are good ones but the absoluteness of his argument makes them kind of meaningless. My general take is that theres a lot of negativity based on real ...

5 2 43

Will Stancil @whstancil.bsky.social · 10h

No one denies this. Other major issues included crime and immigration. What all of these things have in common is they got talked about a ton in the media but were at fairly normal levels in real life. A clue!

bfg123.bsky.social @bfg123.bsky.social · 10h

Most polling aligns on inflation / cost of living being the number 1 issue in 24 actually. Strange to lie about this!

Will Stancil @whstancil.bsky.social · 10h

People are asking why I’m so insistent that inflation doesn’t matter, but it’s not that I think it doesn’t, it’s that I think the null hypothesis is probably correct until demonstrated otherwise, and when the alternative hypothesis must be aggressively p-hacked to survive, I become skeptical of it

9 5 74

Will Stancil @whstancil.bsky.social · 10h

Well we all know it’s better to have wrong explanations with a veneer of empiricism instead of right explanations that are difficult to empirically measure

3 2

Will Stancil @whstancil.bsky.social · 10h

that’s nonresponsive

1 1

Will Stancil @whstancil.bsky.social · 10h

People are asking why I’m so insistent that inflation doesn’t matter, but it’s not that I think it doesn’t, it’s that I think the null hypothesis is probably correct until demonstrated otherwise, and when the alternative hypothesis must be aggressively p-hacked to survive, I become skeptical of it

4 2 35

Will Stancil @whstancil.bsky.social · 10h

“Only during general elections, not midterms. Only among low-info voters. Only when it follows a long period of low inflation. Only nominal prices matter, not real prices.” This is p-hacking. I promise that’s what you are doing

2 1 51

Will Stancil @whstancil.bsky.social · 11h

Basically, what a lot of folks are doing is asserting “Elections are driven by inflation and cost of living” and then adding a thousand caveats until all the obvious rebuttals and counterexamples go away. And that is literally the definition of p-hacking and it’s extremely terrible analytic practice

10 9 120

Will Stancil @whstancil.bsky.social · 11h

If inflation had a predictable effect on political opinion you should be able to model that effect in a way that makes useful predictions, and yet you cannot, because it’s wrong

2 3

Will Stancil @whstancil.bsky.social · 11h

“There’s a magic class of voters who only showed up in 2024 and who care only about inflation even though inflation had been pretty low for almost two years and all of this is kind of a one-off thing that fails to match any previous trend” is just, like, not a very good explanation

7 1 11