Andrew Little

It is in fact time for some game theory

What would you say if you saw this in another country?

Yeah others may have better workflow than me but I'm generally just running Claude code from the terminal. If starting something new just make sure to specify to write in R. If working in an existing project it (in my experience) just works in whatever language it sees there.

Yeah I’d say they are good at narrow tasks (checking derivations, etc) hit or miss at somewhat broader tasks (complete this proof, extend the analysis with a different assumption), basically hopeless at ”write me a model of X”

Either of those better than “theorist who started writing empirical papers after tenure”

Maybe add a tuning parameter between our messy code and good general style?

Neat idea, did you have it export a md file for future use or some similar trick?

Appears there is an information theory result which says this more precisely (average entropy decreases when conditioning) but not sure it has a catchy name

Law of total variance says on average you should be less surprised over time (sort of)

That’s a more plausible if less fun theory 😉

Is this some kind of 3D horseshoe theory where extreme anti-elite/anti-establishment people find common cause with… the elite establishment?

☀️☀️ Exciting news in Madrid!

We are VERY happy to officially announce the creation of the Instituto Juan Linz (aka “IJL”).
ijlinz.es

The IJL ("Juan Linz Institute" in English) is dedicated to cutting-edge research and teaching in political science, sociology, economic history and more.
thread 1/12

The news about the Washington Post is terrible in many ways but it's weird to see this cast as "capitulating to Trump" when they are keeping (and at least claiming to focus more) on national/political news which has done a lot of good reporting criticizing him.

1. Trump certainly wants to fully rig elections.
2. He has no actual plausible plan to do so.
3. We should absolutely take the threat very seriously and actively do what's possible to protect elections and democracy.
4. What's actually likely is success or failure at the margins; every bit counts.

As usual Anne is clearer than me 😂. My gripe is not about the speed but the claim that the total amount of backsliding in the US is coded as just a bit less than Venezuela over the entire period where elections blatantly stolen 10s of thousands killed, etc.

Let’s maybe ask Dorothy to adjudicate this one

Seriously?

Started, going

One can simply go to www.justice.gov/epstein and search “political science“

Also to nitpick the idea that the amount of backsliding experienced in the US in the past year is even in the ballpark of the sum total of what happened in Venezuela doesn't really pass a smell test.

Yeah can definitely cause some damage. I'm just still a committed minimalist definer of democracy (maybe the last one?!?) so tend to view things through the lens of "will this help the incumbent stay in power"

I think he just put it together. Another wonkier issue is I think a lot of the chosen dimensions were inspired by the US case.

My views are closest to this paragraph from the end of the article. To pick a concrete example, Trump trying to force his IRS to pay him billions is clearly the behavior of someone trying to take what they can now rather than consolidating his party's power.

What we see under Trump instead is a bunch of short term/unilateral action. Which can be very bad but if anything could undermine Rs ability to win elections in 2026/2028.

The central force leading to major and durable backsliding in the comparison cases was the weakening/destruction of fair electorla competition. In the index this gets equal weight as central bank independence (important but not in any convention democracy definition/measure I know of).

How steep is Trump’s democratic backsliding? The erosion of established norms has been dramatic but institutions are holding up

Quick thought on the John Murdoch-Burns piece in FT: it's an interesting exercise with some strong insights, but the headline result most are pointing to (US Backsliding faster than Hungary/Venezuela/Russia!) is kind of misleading.

www.ft.com/content/b474...

Yeah based on what I read certainly some value in the individual measures but aggregation is another story

Haven‘t read carefully yet (Anne was the commenter though I agree with what she said) but do you think “the US backslid as much in 1 year as Venezuela did in 12” meets a face validity check?

new frontiers in maga: italian man tries to convince other people that actually scandinavians are an inferior race

A good way to make your academic life more annoying is to ask it to write negative referee reports for your working papers (prob worth it though)